Title: Brewing science and practice
Title: Brewing science and practice
Title: Brewing science and practice
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
<strong>Brewing</strong><br />
Science <strong>and</strong> <strong>practice</strong><br />
Dennis E. Briggs, Chris A. Boulton, Peter A. Brookes <strong>and</strong><br />
Roger Stevens<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Published by Woodhead Publishing Limited, Abington Hall, Abington<br />
Cambridge CB1 6AH, Engl<strong>and</strong><br />
www.woodhead-publishing.com<br />
Published in North America by CRC Press LLC, 2000 Corporate Blvd, NW<br />
Boca Raton FL 33431, USA<br />
First published 2004, Woodhead Publishing Limited <strong>and</strong> CRC Press LLC<br />
ß2004, Dennis E. Briggs, Chris A. Boulton, Peter A. Brookes <strong>and</strong> Roger Stevens<br />
The authors have asserted their moral rights.<br />
This book contains information obtained from authentic <strong>and</strong> highly regarded sources. Reprinted<br />
material is quoted with permission, <strong>and</strong> sources are indicated. Reasonable efforts have been<br />
made to publish reliable data <strong>and</strong> information, but the authors <strong>and</strong> the publishers cannot assume<br />
responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone<br />
else associated with this publication, shall be liable for any loss, damage or liability directly or<br />
indirectly caused or alleged to be caused by this book.<br />
Neither this book nor any part may be reproduced or transmitted in any form or by any<br />
means, electronic or mechanical, including photocopying, microfilming, <strong>and</strong> recording, or by<br />
any information storage or retrieval system, without prior permission in writing from the<br />
publishers.<br />
The consent of Woodhead Publishing Limited <strong>and</strong> CRC Press LLC does not extend to<br />
copying for general distribution, for promotion, for creating new works, or for resale. Specific<br />
permission must be obtained in writing from Woodhead Publishing Limited or CRC Press LLC<br />
for such copying.<br />
Trademark notice: Product or corporate names may be trademarks or registered trademarks, <strong>and</strong><br />
are used only for identification <strong>and</strong> explanation, without intent to infringe.<br />
British Library Cataloguing in Publication Data<br />
Acatalogue record for this book is available from the British Library.<br />
Library of Congress Cataloging-in-Publication Data<br />
Acatalog record for this book is available from the Library of Congress.<br />
Woodhead Publishing Limited ISBN 185573 490 7(book) 185573 906 2(e-book)<br />
CRC Press ISBN 0-8493-2547-1<br />
CRC Press order number: WP2547<br />
The publisher's policy is to use permanent paper from mills that operate asustainable forestry<br />
policy, <strong>and</strong> which have been manufactured from pulp which is processed using acid-free <strong>and</strong><br />
elementary chlorine-free <strong>practice</strong>s. Furthermore, the publisher ensures that the text paper <strong>and</strong><br />
cover board used have met acceptable environmental accreditation st<strong>and</strong>ards.<br />
Project managed by Macfarlane Production Services, Markyate, St Albans, Hertfordshire<br />
(e-mail: macfarl@aol.com)<br />
Typeset by MHL Typesetting Ltd, Coventry, Warwickshire<br />
Printed by TJ International, Cornwall, Engl<strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Contents<br />
Preface<br />
1 An outline of brewing<br />
1.1 Introduction<br />
1.2 Malts<br />
1.3 Mash tun adjuncts<br />
1.4 <strong>Brewing</strong> liquor<br />
1.5 Milling <strong>and</strong> mashing in<br />
1.6 Mashing <strong>and</strong> wort separation systems<br />
1.7 The hop-boil <strong>and</strong> copper adjuncts<br />
1.8 Wort clarification, cooling <strong>and</strong> aeration<br />
1.9 Fermentation<br />
1.10 The processing of beer<br />
1.11 Types of beer<br />
1.12 Analytical systems<br />
1.13 The economics of brewing<br />
1.14 Excise<br />
1.15 References <strong>and</strong> further reading<br />
1.15.1 The systems of malting <strong>and</strong> brewing analysis<br />
1.15.2 General references<br />
2 Malts, adjuncts <strong>and</strong> supplementary enzymes<br />
2.1 Grists <strong>and</strong> other sources of extract<br />
2.2 Malting<br />
2.2.1 Malting in outline<br />
2.2.2 Changes occurring in malting grain<br />
2.2.3 Malting technology<br />
2.2.4 Malt analyses<br />
2.2.5 Types of kilned malt<br />
2.2.6 Special malts<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
vi Contents<br />
2.2.7 Malt specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32<br />
2.3 Adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34<br />
2.3.1 Mash tun adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34<br />
2.3.2 Copper adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40<br />
2.4 Priming sugars, caramels, malt colourants <strong>and</strong> Farbebier . . . . . . . . . . 45<br />
2.5 Supplementary enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46<br />
2.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50<br />
3 Water, effluents <strong>and</strong> wastes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52<br />
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52<br />
3.2 Sources of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53<br />
3.3 Preliminary water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57<br />
3.4 Secondary water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60<br />
3.5 Grades of water used in breweries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64<br />
3.6 The effects of ions on the brewing process . . . . . . . . . . . . . . . . . . . . . . . . . 65<br />
3.7 Brewery effluents, wastes <strong>and</strong> by-products . . . . . . . . . . . . . . . . . . . . . . . . . 68<br />
3.7.1 The characterization of waste water . . . . . . . . . . . . . . . . . . . . . . . . 69<br />
3.7.2 The characteristics of some brewery wastes <strong>and</strong><br />
by-products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71<br />
3.8 The disposal of brewery effluents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73<br />
3.8.1 Preliminary treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73<br />
3.8.2 Aerobic treatments of brewery effluents . . . . . . . . . . . . . . . . . . . 75<br />
3.8.3 Sludge treatments <strong>and</strong> disposal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78<br />
3.8.4 Anaerobic <strong>and</strong> mixed treatments of brewery effluents . . . . . 79<br />
3.9 Other water treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82<br />
3.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82<br />
4 The <strong>science</strong> of mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85<br />
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85<br />
4.2 Mashing schedules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88<br />
4.3 Altering mashing conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95<br />
4.3.1 The grist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95<br />
4.3.2 Malts in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97<br />
4.3.3 Mashing with adjuncts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101<br />
4.3.4 The influence of mashing temperatures <strong>and</strong> times on<br />
wort quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104<br />
4.3.5 Non-malt enzymes in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110<br />
4.3.6 Mashing liquor <strong>and</strong> mash pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113<br />
4.3.7 Mash thickness, extract yield <strong>and</strong> wort quality . . . . . . . . . . . . . 116<br />
4.3.8 Wort separation <strong>and</strong> sparging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119<br />
4.4 Mashing biochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122<br />
4.4.1 Wort carbohydrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122<br />
4.4.2 Starch degradation in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127<br />
4.4.3 Non-starch polysaccharides in mashing . . . . . . . . . . . . . . . . . . . . 136<br />
4.4.4 Proteins, peptides <strong>and</strong> amino acids . . . . . . . . . . . . . . . . . . . . . . . . . 142<br />
4.4.5 Nucleic acids <strong>and</strong> related substances . . . . . . . . . . . . . . . . . . . . . . . 146<br />
4.4.6 Miscellaneous substances containing nitrogen . . . . . . . . . . . . . . 146<br />
4.4.7 Vitamins <strong>and</strong> yeast growth factors . . . . . . . . . . . . . . . . . . . . . . . . . 149<br />
4.4.8 Lipids in mashing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
4.4.9 Phenols<br />
4.4.10 Miscellaneous acids<br />
4.4.11 Inorganic ions in sweet wort<br />
4.5 Mashing <strong>and</strong> beer flavour<br />
4.6 Spent grains<br />
4.7 References<br />
5 The preparation of grists<br />
5.1 Intake, h<strong>and</strong>ling <strong>and</strong> storage of raw materials<br />
5.2 The principles of milling<br />
5.3 Laboratory mills<br />
5.4 Dry roller milling<br />
5.5 Impact mills<br />
5.6 Conditioned dry milling<br />
5.7 Spray steep roller milling<br />
5.8 Steep conditioning<br />
5.9 Milling under water<br />
5.10 Grist cases<br />
5.11 References<br />
6 Mashing technology<br />
6.1 Introduction<br />
6.2 Mashing in<br />
6.3 The mash tun<br />
6.3.1 Construction<br />
6.3.2 Mash tun operations<br />
6.4 Mashing vessels for decoction, double mashing <strong>and</strong> temperatureprogrammed<br />
infusion mashing systems<br />
6.4.1 Decoction <strong>and</strong> double mashing<br />
6.4.2 Temperature-programmed infusion mashing<br />
6.5 Lauter tuns<br />
6.6 The Strainmaster<br />
6.7 Mash filters<br />
6.8 The choice of mashing <strong>and</strong> wort separation systems<br />
6.9 Other methods of wort separation <strong>and</strong> mashing<br />
6.10 Spent grains<br />
6.11 Theory of wort separation<br />
6.12 References<br />
7 Hops<br />
7.1 Introduction<br />
7.2 Botany<br />
7.3 Cultivation<br />
7.4 Drying<br />
7.5 Hop products<br />
7.5.1 Hop pellets<br />
7.5.2 Hop extracts<br />
7.5.3 Hop oils<br />
7.6 Pests <strong>and</strong> diseases<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
7.6.1 Damson-hop aphid<br />
7.6.2 (Red) Spider Mite<br />
7.6.3 Other pests<br />
7.6.4 Downy mildew<br />
7.6.5 Powdery mildew<br />
7.6.6 Verticillium wilt<br />
7.6.7 Virus diseases<br />
7.7 Hop varieties<br />
7.8 References<br />
8 The chemistry of hop constituents<br />
8.1 Introduction<br />
8.2 Hop resins<br />
8.2.1 Introduction<br />
8.2.2 Biosynthesis of the hop resins<br />
8.2.3 Analysis of the hop resins<br />
8.2.4 Isomerization of the -acids<br />
8.2.5 Hard resins <strong>and</strong> prenylflavonoids<br />
8.2.6 Oxidation of the hop resins<br />
8.3 Hop oil<br />
8.3.1 Introduction<br />
8.3.2 Hydrocarbons<br />
8.3.3 Oxygen-containing components<br />
8.3.4 Sulphur-containing compounds<br />
8.3.5 Most potent odorants in hop oil<br />
8.3.6 Hop oil constituents in beer<br />
8.3.7 Post fermentation aroma products<br />
8.4 Hop polyphenols (tannins)<br />
8.5 Chemical identification of hop cultivars<br />
8.6 References<br />
9 Chemistry of wort boiling<br />
9.1 Introduction<br />
9.2 Carbohydrates<br />
9.3 Nitrogenous constituents<br />
9.3.1 Introduction<br />
9.3.2 Proteins<br />
9.4 Carbohydrate-nitrogenous constituent interactions<br />
9.4.1 Melanoidins<br />
9.4.2 Caramel<br />
9.5 Protein-polyphenol (tannin) interactions<br />
9.6 Copper finings <strong>and</strong> trub formation<br />
9.7 References<br />
10 Wort boiling, clarification, cooling <strong>and</strong> aeration<br />
10.1 Introduction<br />
10.2 The principles of heating wort<br />
10.3 Types of coppers<br />
10.4 The addition of hops<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
10.5 Pressurized hop-boiling systems<br />
10.5.1 Low-pressure boiling<br />
10.5.2 Dynamic low-pressure boiling<br />
10.5.3 Continuous high-pressure boiling<br />
10.6 The control of volatile substances in wort<br />
10.7 Energy conservation <strong>and</strong> the hop-boil<br />
10.8 Hot wort clarification<br />
10.9 Wort cooling<br />
10.10 The cold break<br />
10.11 Wort aeration/oxygenation<br />
10.12 References<br />
11 Yeast biology<br />
11.1 Historical note<br />
11.2 Taxonomy<br />
11.3 Yeast ecology<br />
11.4 Cellular composition<br />
11.5 Yeast morphology<br />
11.6 Yeast cytology<br />
11.6.1 Cell wall<br />
11.6.2 The periplasm<br />
11.6.3 The plasma membrane<br />
11.6.4 The cytoplasm<br />
11.6.5 Vacuoles <strong>and</strong> intracellular membrane systems<br />
11.6.6 Mitochondria<br />
11.6.7 The nucleus<br />
11.7 Yeast cell cycle<br />
11.7.1 Yeast sexual cycle<br />
11.8 Yeast genetics<br />
11.8.1 Methods of genetic analysis<br />
11.8.2 The yeast genome<br />
11.9 Strain improvement<br />
11.10 References<br />
12 Metabolism of wort by yeast<br />
12.1 Introduction<br />
12.2 Yeast metabolism ±an overview<br />
12.3 Yeast nutrition<br />
12.3.1 Water relations<br />
12.3.2 Sources of carbon<br />
12.3.3 Sources of nitrogen<br />
12.3.4 Sources of minerals<br />
12.3.5 Growth factors<br />
12.4 Nutrient uptake<br />
12.4.1 Sugar uptake<br />
12.4.2 Uptake of nitrogenous nutrients<br />
12.4.3 Lipid uptake<br />
12.4.4 Ion uptake<br />
12.4.5 Transport of the products of fermentation<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
12.5 Sugar metabolism<br />
12.5.1 Glycolysis<br />
12.5.2 Hexose monophosphate (pentose phosphate) pathway<br />
12.5.3 Tricarboxylic acid cycle<br />
12.5.4 Electron transport <strong>and</strong> oxidative phosphorylation<br />
12.5.5 Fermentative sugar catabolism<br />
12.5.6 Gluconeogenesis <strong>and</strong> the glyoxylate cycle<br />
12.5.7 Storage carbohydrates<br />
12.5.8 Regulation of sugar metabolism<br />
12.5.9 Ethanol toxicity <strong>and</strong> tolerance<br />
12.6 The role of oxygen<br />
12.7 Lipid metabolism<br />
12.7.1 Fatty acid metabolism<br />
12.7.2 Phospholipids<br />
12.7.3 Sterols<br />
12.8 Nitrogen metabolism<br />
12.9 Yeast stress responses<br />
12.10 Minor products of metabolism contributing to beer flavour<br />
12.10.1 Organic <strong>and</strong> fatty acids<br />
12.10.2 Carbonyl compounds<br />
12.10.3 Higher alcohols<br />
12.10.4 Esters<br />
12.10.5 Sulphur-containing compounds<br />
12.11 References<br />
13 Yeast growth<br />
13.1 Introduction<br />
13.2 Measurement of yeast biomass<br />
13.3 Batch culture<br />
13.3.1 Brewery batch fermentations<br />
13.3.2 Effects of process variables on fermentation performance<br />
13.4 Yeast ageing<br />
13.5 Yeast propagation<br />
13.5.1 Maintenance <strong>and</strong> supply of yeast cultures<br />
13.5.2 Laboratory yeast propagation<br />
13.5.3 Brewery propagation<br />
13.6 Fed-batch cultures<br />
13.7 Continuous culture<br />
13.8 Immobilized yeast reactors<br />
13.9 Growth on solid media<br />
13.10 Yeast identification<br />
13.10.1 Microbiological tests<br />
13.10.2 Biochemical tests<br />
13.10.3 Tests based on cell surface properties<br />
13.10.4 Non-traditional methods<br />
13.11 Measurement of viability<br />
13.12 Assessment of yeast physiological state<br />
13.13 References<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
14 Fermentation technologies<br />
14.1 Introduction<br />
14.2 Basic principles of fermentation technology<br />
14.2.1 Fermentability of wort<br />
14.2.2 Time course of fermentation<br />
14.2.3 Heat output in fermentation<br />
14.3 Bottom fermentation systems<br />
14.3.1 Choice, size <strong>and</strong> shape of vessels<br />
14.3.2 Construction of cylindroconical vessels<br />
14.3.3 Operation of cylindroconical vessels<br />
14.4 Top fermentation systems<br />
14.4.1 Traditional top fermentation<br />
14.4.2 Yorkshire square fermentation<br />
14.4.3 Burton Union fermentation<br />
14.5 Continuous fermentation<br />
14.5.1 Early systems of continuous fermentation<br />
14.5.2 The New Zeal<strong>and</strong> system<br />
14.5.3 Continuous primary fermentation with immobilized yeast<br />
14.6 Fermentation control systems<br />
14.6.1 Specific gravity changes<br />
14.6.2 Other methods<br />
14.7 Summary<br />
14.8 References<br />
15 Beer maturation <strong>and</strong> treatments<br />
15.1 Introduction<br />
15.2 Maturation: flavour <strong>and</strong> aroma changes<br />
15.2.1 Principles of secondary fermentation<br />
15.2.2 Important flavour changes 5<br />
15.2.3 Techniques of maturation<br />
15.2.4 Flavour, aroma <strong>and</strong> colour adjustments by addition<br />
15.2.5 Maturation vessels<br />
15.3 Stabilization against non-biological haze<br />
15.3.1 Mechanisms for haze formation<br />
15.3.2 Removal of protein<br />
15.3.3 Removal of polyphenols<br />
15.3.4 Combined treatments to remove protein <strong>and</strong> polyphenols<br />
15.3.5 Hazes from other than protein or polyphenols<br />
15.4 Carbonation<br />
15.4.1 Carbon dioxide saturation<br />
15.4.2 Carbon dioxide addition<br />
15.4.3 Carbon dioxide recovery<br />
15.5 Clarification <strong>and</strong> filtration<br />
15.5.1 Removal of yeast <strong>and</strong> beer recovery<br />
15.5.2 Beer filtration<br />
15.6 Special beer treatments<br />
15.6.1 Low-alcohol <strong>and</strong> alcohol-free beers<br />
15.6.2 Ice beers<br />
15.6.3 Diet beers<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
15.7 Summary<br />
15.8 References<br />
16 Native African beers<br />
16.1 Introduction<br />
16.1.1 An outline of the stages of production<br />
16.1.2 Bouza<br />
16.1.3 Merissa<br />
16.1.4 Busaa <strong>and</strong> some other beers<br />
16.1.5 Southern African beers<br />
16.2 Malting sorghum <strong>and</strong> millets<br />
16.3 <strong>Brewing</strong> African beers on an industrial scale<br />
16.4 Attempts to obtain stable African beers<br />
16.5 Beer composition <strong>and</strong> its nutritional value<br />
16.6 References<br />
17 Microbiology<br />
17.1 Introduction<br />
17.2 The microbiological threat to the brewing process<br />
17.3 Beer spoilage micro-organisms<br />
17.3.1 Detection of brewery microbial contaminants<br />
17.3.2 Identification of brewery bacteria<br />
17.3.3 Gram negative beer spoiling bacteria<br />
17.3.4 Gram positive beer spoiling bacteria<br />
17.3.5 Beer spoilage yeasts<br />
17.3.6 Microbiological media <strong>and</strong> the cultivation of<br />
micro-organisms<br />
17.4 Microbiological quality assurance<br />
17.5 Sampling<br />
17.5.1 Sampling devices<br />
17.6 Disinfection of pitching yeast<br />
17.7 Cleaning in the brewery<br />
17.7.1 Range of cleaning operations<br />
17.7.2 CIP systems<br />
17.7.3 Cleaning agents<br />
17.7.4 Cleaning beer dispense lines<br />
17.7.5 Validation of CIP<br />
17.8 References<br />
18 Brewhouses: types, control <strong>and</strong> economy<br />
18.1 Introduction<br />
18.2 History of brewhouse development<br />
18.2.1 The tower brewery lay-out<br />
18.2.2 The horizontal brewery lay-out<br />
18.3 Types of modern brewhouses<br />
18.3.1 Experimental brewhouses<br />
18.3.2 Micro- <strong>and</strong> pub breweries<br />
18.4 Control of brewhouse operations<br />
18.4.1 Automation in the brewhouse<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
18.4.2 Scheduling of brewhouse operations<br />
18.5 Economic aspects of brewhouses<br />
18.6 Summary<br />
18.7 References<br />
19 Chemical <strong>and</strong> physical properties of beer<br />
19.1 Chemical composition of beer<br />
19.1.1 Inorganic constituents<br />
19.1.2 Alcohol <strong>and</strong> original extract<br />
19.1.3 Carbohydrates<br />
19.1.4 Other constituents containing carbon, hydrogen <strong>and</strong><br />
oxygen<br />
19.1.5 Nitrogenous constituents<br />
19.1.6 Sulphur-containing constituents<br />
19.2 Nutritive value of beer<br />
19.3 Colour of beer<br />
19.4 Haze<br />
19.4.1 Measurement of haze<br />
19.4.2 Composition <strong>and</strong> formation of haze<br />
19.4.3 Prediction of haze <strong>and</strong> beer stability<br />
19.4.4 Practical methods for improving beer stability<br />
19.5 Viscosity<br />
19.6 Foam characteristics <strong>and</strong> head retention<br />
19.6.1 Methods of assessing foam characteristics<br />
19.6.2 Beer components influencing head retention<br />
19.6.3 Head retention <strong>and</strong> the brewing process<br />
19.7 Gushing<br />
19.8 References<br />
20 Beer flavour <strong>and</strong> sensory assessment<br />
20.1 Introduction<br />
20.2 Flavour ±taste <strong>and</strong> odour<br />
20.3 Flavour stability<br />
20.4 Sensory analysis<br />
20.5 References<br />
21 Packaging<br />
21.1 Introduction<br />
21.2 General overview of packaging operations<br />
21.3 Bottling<br />
21.3.1 Managing the bottle flow<br />
21.3.2 Managing the beer flow<br />
21.3.3 Managing plant cleaning<br />
21.3.4 Materials for making bottles<br />
21.4 Canning<br />
21.4.1 The beer can<br />
21.4.2 Preparing cans at the brewery for filling<br />
21.4.3 Can filling<br />
21.4.4 Can closing (seaming)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
21.4.5 Widgets in cans<br />
21.5 Kegging<br />
21.5.1 The keg<br />
21.5.2 Treatment of beer for kegging<br />
21.5.3 H<strong>and</strong>ling of kegs<br />
21.5.4 Keg internal cleaning <strong>and</strong> filling<br />
21.5.5 Keg capping <strong>and</strong> labelling<br />
21.5.6 Smooth flow ale in kegs<br />
21.6 Cask beer<br />
21.6.1 The cask<br />
21.6.2 H<strong>and</strong>ling casks<br />
21.6.3 Preparing beer for cask filling<br />
21.6.4 Cask filling<br />
21.7 Summary<br />
21.8 References<br />
22 Storage <strong>and</strong> distribution<br />
22.1 Introduction<br />
22.2 Warehousing<br />
22.2.1 Principles of warehouse operation<br />
22.2.2 Safety in the warehouse<br />
22.3 Distribution<br />
22.3.1 Logistics<br />
22.3.2 Quality assurance<br />
22.4 Summary<br />
22.5 References<br />
23 Beer in the trade<br />
23.1 Introduction<br />
23.2 History<br />
23.3 Beer cellars<br />
23.3.1 Hygiene<br />
23.3.2 Temperature<br />
23.3.3 Lighting<br />
23.4 Beer dispense<br />
23.4.1 Keg beer<br />
23.4.2 Cask beer<br />
23.4.3 Bottled <strong>and</strong> canned beer<br />
23.5 Quality control<br />
23.6 New developments in trade quality<br />
23.7 Summary<br />
23.8 References<br />
Appendix: units <strong>and</strong> some data of use in brewing<br />
Table A1 SI derived units<br />
Table A2 Prefixes for SI units<br />
Table A3 Comparison of thermometer scales<br />
Table A4 Interconversion factors for units of measurement<br />
Table A5 Specific gravity <strong>and</strong> extract table<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A6 Equivalence between Institute of <strong>Brewing</strong> units of hot<br />
water extract<br />
Table A7 Solution divisors of some sugars<br />
Table A8 Some properties of water at various temperatures<br />
Table A9 The density <strong>and</strong> viscosity of water at various temperatures<br />
Table A10 Some more properties of water<br />
Table A11 The relationship between the absolute pressure <strong>and</strong> the<br />
temperature of water-saturated steam<br />
Table A12 The solubility of pure gases in water at different<br />
temperatures<br />
Table A13 Salts in brewing liquors<br />
Table A14 Units of degrees of water hardness<br />
Table A15 Characteristics of some brewing materials<br />
Table A16 Pasteurization units<br />
Fig. A1 The relationships between ethanol/water mixtures <strong>and</strong> the<br />
densities of the solutions.<br />
References<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Preface<br />
ThetwovolumesofthesecondeditionofMalting<strong>and</strong><strong>Brewing</strong>ScienceI,Malt<strong>and</strong>Sweet<br />
Wort <strong>and</strong> II, Hopped Wort <strong>and</strong> Beer, by James S. Hough, Dennis E. Briggs, Roger<br />
Stevens <strong>and</strong> Tom W. Young, appeared in 1981 <strong>and</strong> 1982. This book provided the<br />
framework for the M.Sc. in Malting <strong>and</strong> <strong>Brewing</strong> Science, the course that was offered by<br />
the British School of Malting <strong>and</strong> <strong>Brewing</strong> in the University of Birmingham (UK). It also<br />
provided the backbone of many other courses. After more than 20 years the dem<strong>and</strong> for<br />
these volumes has continued, although they are increasingly out of date. Malts <strong>and</strong><br />
Malting, by Dennis E. Briggs, appeared in 1998, <strong>and</strong> <strong>Brewing</strong> Yeast <strong>and</strong> Fermentation,<br />
by Chris Boulton <strong>and</strong> David Quain, became available in 2001. These books cover their<br />
named topics in depth. However, the need for an up-to-date, integrated textbook on<br />
brewing, comparable in scope <strong>and</strong> depth of coverage to Malting <strong>and</strong> <strong>Brewing</strong> Science,<br />
remained.<br />
<strong>Brewing</strong>:Science<strong>and</strong><strong>practice</strong>isintendedtomeetthisneed.Decidingonthedetailsof<br />
the coverage has given rise to some anxious discussions. Practically it is impossible to<br />
describe all aspects of all the varieties of brewing processes in depth, in one moderately<br />
sized volume. Inevitably it has been necessary to assume some background knowledge of<br />
physics, chemistry, biology, <strong>and</strong> engineering. However, the book is underst<strong>and</strong>able to<br />
people without detailed knowledge in these areas. The references at the end of each<br />
chapter provide guidance for further reading. Since the wide range of kinds of brewing<br />
operations, from simple, low-volume, single-line breweries to extremely large, highly<br />
complex, multiple-line installations, does not allow a single description of brewing<br />
activities, the book concentrates on the principles of the various brewing processes.<br />
<strong>Brewing</strong> is carried out all over the world <strong>and</strong>, unsurprisingly, different terminologies<br />
<strong>and</strong> methods of measurement <strong>and</strong> analysis are used. The different systems of units <strong>and</strong><br />
analyses are explained in the text <strong>and</strong> conversion factors (where valid) <strong>and</strong> some other<br />
useful data are given in the Appendix. Alist of abbreviations is included in the index for<br />
reference. The index also includes a list of formulae<br />
First of all the authors warmly thank our wives, Rosemary, Wendy, Stella <strong>and</strong> Betty,<br />
for their unfailing patience <strong>and</strong> good-humoured support. We have also been given a great<br />
deal of help from our colleagues <strong>and</strong> friends. We are grateful to Mrs Doreen Hough for<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
permission to use some of the late Professor Jim Hough's diagrams. Permission to use<br />
other diagrams is acknowledged in the text. We would like to thank: Mrs Marjorie<br />
Anderson, Dr John M. H. Andrews, Mrs Marjorie Anderson, Dr Raymond G. Anderson,<br />
Mr David J. Banfield, Mr Zane C. Barnes, Herr Volker Borngraber, Mr Andy Carter, Dr<br />
Peter Darby, Mr J. Brian Eaton, Dr David L. Griggs, Dr Paul K. Hegarty, Mrs Sue M.<br />
Henderson, Mr James Johnstone, Mr Roy F. Lindsay, Dr G. C. Linsley-Noakes, Dr David<br />
E. Long, Mr John MacDonald, Dr Ray Marriott, Mr P. A. (Tom) Martin, Dr A. Peter<br />
Maule, Ms Elaine McCrimmon, Dr Philip Morrall, Dr Ray Neve, Dr George Philliskirk,<br />
Dr David E. Quain, Mr Trevor R. Roberts, Mr Derek Wareham <strong>and</strong> Dr Richard D. J.<br />
Webster. We also wish to thank Coors Brewers for the use of the Technical Centre,<br />
Burton-on-Trent.<br />
We apologise if any acknowledgements have been omitted.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1<br />
An outline of brewing<br />
1.1 Introduction<br />
Beers<strong>and</strong>beer-likebeveragesmaybepreparedfromrawcerealgrains,maltedcerealgrains<br />
<strong>and</strong> (historically) bread. This book is primarily concerned with beers of the types that<br />
originated in Europe, but which are now produced world-wide. However, an account is<br />
given of `African-style' beers (Chapter 16). The most simple preparation of European-style<br />
beers involves (a) incubating <strong>and</strong> extracting malted, ground up cereal grains (usually<br />
barley) with warm water. Sometimes the ground malt is mixed with other starchy materials<br />
<strong>and</strong>/or enzymes. (b) The solution obtained is boiled with hops or hop preparations. (c) The<br />
boiled solution is clarified <strong>and</strong> cooled. (d) The cooled liquid is fermented by added yeast.<br />
Usually the beer is clarified, packaged <strong>and</strong> served while effervescent with escaping carbon<br />
dioxide. In this chapter the preparation of beers is outlined <strong>and</strong> the brewers' vocabulary is<br />
introduced. Beers are made in amounts ranging from afew hectolitres (hl) aweek to<br />
thous<strong>and</strong>s of hl. They are made using various different systems of brewing.<br />
1.2 Malts<br />
Malts are made from selected cereal grain, usually barley, (but sometimes wheat, rye,<br />
oats, sorghum or millet), that has been cleaned <strong>and</strong> stored until dormancy has declined<br />
<strong>and</strong> it is needed. It is then germinated under controlled conditions. Their preparation is<br />
outlined in Chapter 2, <strong>and</strong> is described in detail in Briggs (1998). The grain is hydrated,<br />
or `steeped', by immersion in water. During steeping the water will be changed at least<br />
once, air may be sucked through the grain during `dry' periods between immersions, <strong>and</strong><br />
may be blown into the grain while immersed. After steeping the grain is drained <strong>and</strong> is<br />
germinated to a limited extent in a cool, moist atmosphere with occasional turning <strong>and</strong><br />
mixing to prevent the rootlets matting together. During germination the acrospire<br />
(coleoptile) grows beneath the husk <strong>and</strong> rootlets grow from the end of the grain, enzymes<br />
accumulate <strong>and</strong> so do sugars <strong>and</strong> other soluble materials. The dead storage tissue of the<br />
grain, the starchy endosperm, is partly degraded, or `modified', <strong>and</strong> its physical strength<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
is reduced. When germination <strong>and</strong> `modification' are sufficiently advanced they are<br />
stopped by kilning. The `green malt' (green in the sense of immature, it is not green in<br />
colour) is kilned, that is, it is dried <strong>and</strong> lightly cooked, or cured, in acurrent of warm to<br />
hot air. Pale, `white' malts are kilned using low temperatures <strong>and</strong> in these enzyme<br />
survival is considerable. In darker, coloured malts, kilned using higher temperatures,<br />
enzyme survival is less. In extreme cases the darkest, special malts are heated in a<br />
roasting drum <strong>and</strong> contain no active enzymes. After kilning the malt is cooled <strong>and</strong><br />
`dressed', that is, the brittle rootlets (`culms', sprouts) are broken off <strong>and</strong> they <strong>and</strong> dust<br />
are removed. The culms are usually used for cattle food. Pale malts are usually stored for<br />
some weeks before use. In contrast to the tough, ungerminated barley grain malt is<br />
`friable', that is, it is easily crushed.<br />
1.3 Mash tun adjuncts<br />
Mash tun adjuncts are preparations of cereals (e.g., flaked maize or rice flakes, wheat<br />
flour, micronized wheat grains, or rice or maize grits which have to be cooked separately<br />
inthebrewery)which maybemixed withground maltinthemashingprocess. Theuseof<br />
an adjunct alters the character of the beer produced. An adjunct's starch is hydrolysed<br />
during mashing by enzymes from the malt, so providing a(sometimes) less expensive<br />
source of sugars as well as changing the character of the wort. Sometimes microbial<br />
enzymes are added to the mash. In afew countries the use of adjuncts is forbidden. In<br />
GermanytheReinheitsgebotstipulatesthatbeermaybemadeonlywithwater,malt,hops<br />
<strong>and</strong> yeast.<br />
1.4 <strong>Brewing</strong> liquor<br />
In brewing, water is commonly known as liquor. It is used for many purposes besides<br />
mashing, including beer dilution at the end of high-gravity brewing, cleaning <strong>and</strong> in<br />
raising steam. Water for each purpose must meet different quality criteria (Chapter 3).<br />
The brewing liquor used in mashing must be essentially `pure', but it must contain<br />
dissolved salts appropriate for the beer being made. The quality of the liquor influences<br />
the character of the beer made from it. Famous brewing locations gained their<br />
reputations, at least in part, from the qualities of the liquors available to them. Thus<br />
Burton-on-Trent is famous for its pale ales, Dublin for its stouts <strong>and</strong> Pilsen for its fine,<br />
pale lagers. It is now usual, at least in larger breweries, to adjust the composition of the<br />
brewing liquor (Chapter 3).<br />
1.5 Milling <strong>and</strong> mashing in<br />
The malt, sometimes premixed with particular adjuncts, is broken up to acontrolled<br />
extent by milling to create the `grist'. The type of mill used <strong>and</strong> the extent to which the<br />
malt (<strong>and</strong> adjunct) is broken down is chosen to suit the types of mashing <strong>and</strong> wortseparation<br />
systems being used (Chapter5).If drymillingisused the grist, possibly mixed<br />
with adjuncts, is collected in a container, the grist case.<br />
At mashing-in (doughing-in) the grist is intimately mixed with brewing liquor, both<br />
flowing at controlled rates, into a mashing vessel at an exactly controlled temperature.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The resulting `mash', with the consistency of athin slurry, is held for aperiod of<br />
`conversion'. The objective is to obtain amash that will yield asuitable `sweet wort', a<br />
liquidrichinmaterialsdissolvedfromthemalt<strong>and</strong>anyadjunctsthathave beenused.The<br />
dissolved material, the `extract', contains soluble substances that were preformed in the<br />
grist <strong>and</strong> other substances (especially carbohydrates derived from starch), that are formed<br />
from previously insoluble materials by enzyme-catalysed hydrolytic breakdown during<br />
mashing.<br />
1.6 Mashing <strong>and</strong> wort separation systems<br />
The major mashing systems are, broadly, (a) the simplest, nearly isothermal, infusion<br />
mashing system, (traditional for British ale brewers); (b) the decoction system,<br />
(traditional for mainl<strong>and</strong> European lager brewers); (c) the double mash system, (common<br />
in North American <strong>practice</strong>); (d) the temperature-programmed infusion mashing system<br />
that is being widely adopted in the UK <strong>and</strong> mainl<strong>and</strong> Europe (Chapters 4<strong>and</strong> 5). Amash<br />
should be held at a chosen temperature (or at successive different temperatures), for predetermined<br />
times, to allow enzymes to `convert' (degrade) the starch <strong>and</strong> dextrins to<br />
soluble sugars, to cause the partial breakdown of proteins, to degrade nucleic acids <strong>and</strong><br />
other substances. At the end of mashing the sweet, or unhopped wort (the solution of<br />
extractives, mainly carbohydrates; the `extract') is separated from the undissolved solids,<br />
the spent grains or draff.<br />
Infusion mashing is carried out in mash tuns. Mash conversion <strong>and</strong> the separation of<br />
the sweet wort from the spent grains take place in this vessel. The coarsely ground grist,<br />
made with a high proportion of well-modified malt, is mashed in to give a relatively<br />
thick, porridge-like mash at 63 67 ëC (145.4 152.6 ëF). After a st<strong>and</strong> of between 30<br />
minutes <strong>and</strong> two <strong>and</strong> a half hours the wort (liquid) is withdrawn from the mash. The first<br />
worts are cloudy <strong>and</strong> are re-circulated, but as the run off is continued the wort becomes<br />
`bright' (clear), because it is filtered through the bed of grist particles. When bright the<br />
wort is either collected in a holding vessel (an underback) or it is moved directly to a<br />
copper to be boiled with hops. Most of the residual extract, initially entrained in the wet<br />
grains, is washed out by sparging (spraying) hot liquor, at 75 80 ëC (167 176 ëF ) over<br />
the goods.<br />
Decoction mashing is carried out with more finely ground grists, originally made with<br />
malts that were undermodified. These mashes are relatively `thin', so they may be moved<br />
by pumping <strong>and</strong> can be stirred. Decoction mashing uses three vessels, a stirred mash<br />
mixing vessel, a stirred decoction vessel or mash cooker <strong>and</strong> a wort separation device,<br />
either a lauter tun or a mash filter. In one traditional mashing programme the grist is<br />
mashed in to give an initial temperature of around 35 ëC (96 ëF). After a st<strong>and</strong> a decoction<br />
is carried out, that is, a proportion of the mash, e.g., a third, is pumped to the mash<br />
cooker, where it is heated to boiling. The boiling mash is pumped back to the mash<br />
mixing vessel <strong>and</strong> is mixed with the vessels contents, raising the temperature to, e.g.,<br />
50 ëC (122 ëF). After another st<strong>and</strong> a second decoction is carried out, increasing the<br />
temperature of the mixed mash to about 65 ëC (149 ëF). A final decoction increases the<br />
mash temperature to about 76 ëC (167 ëF). The mash is then transferred to a lauter tun or a<br />
mash filter. The sweet wort <strong>and</strong> spargings are collected, ready to be boiled with hops.<br />
Typically, double-mashing uses nitrogen- (`protein-') <strong>and</strong> enzyme-rich malts <strong>and</strong><br />
substantial quantities of maize or rice grits. It also involves the use of three vessels. Most<br />
of the malt grist is mashed into a mash-mixing vessel to give a mash at around 38 ëC<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(100.4ëF). The grits, mixed with asmall proportion of ground malt <strong>and</strong>/or apreparation<br />
of microbial enzymes, are mashed in aseparate vessel called acereal cooker. The<br />
contents are carefully heated with mixing, <strong>and</strong> arest at about 70ëC (158ëF), to 100ëC<br />
(212ëF) to disperse the starch <strong>and</strong> partly liquefy it. The adjunct mash is pumped from the<br />
cereal cooker into the malt mash, with continuous mixing, to give afinal temperature of<br />
about 70ëC (158ëF). After ast<strong>and</strong> the mash is heated to about 73ëC (163.4ëF), then it is<br />
usually transferred to alauter tun for wort collection.<br />
Temperature-programmed infusion mashing is increasingly displacing older mashing<br />
systems. The grist is finely ground <strong>and</strong> the mash is made `thin' to allow it to be stirred.<br />
The grist is mashed into astirred <strong>and</strong> externally heated mash-mixing vessel to give an<br />
initial temperature of 35ëC (95ëF) for apoorly modified malt or 50ëC (122ëF), or more,<br />
for abetter modified malt. The mash is heated, with `st<strong>and</strong>s' typically at 50ëC (122ëF),<br />
65ëC (149ëF) <strong>and</strong> 75ëC (167ëC). Then the sweet wort is collected using alauter tun or a<br />
mash filter.<br />
1.7 The hop-boil <strong>and</strong> copper adjuncts<br />
The sweet wort is transferred to avessel, acopper or kettle, in which it is boiled with hops<br />
or hop preparations, usually for 1±2 hours. Hops are the female cones of hop plants. They<br />
may be used whole, or ground up, or as pellets or as extracts. The choice dictates the type<br />
of equipment used inthe next stage of brewing. Pelleted powders are often preferred. Hops<br />
contribute various groups of substances to the wort. During boiling anumber of changes<br />
occurinthewortofwhichthemoreobviousarethecoagulationofproteinas`hotbreak'or<br />
`trub', the gaining of bitterness <strong>and</strong> hop aroma <strong>and</strong> the destruction of micro-organisms<br />
(Chapters9<strong>and</strong>10).Evaporation ofthe wort,reduces the volumeby,say,7±10%, <strong>and</strong>soit<br />
is concentrated. Unwanted flavour-rich <strong>and</strong> aromatic volatile substances are removed.<br />
When used, sugars, syrups <strong>and</strong> even malt extracts (copper adjuncts) are dispersed <strong>and</strong><br />
dissolve in the wort during the copper boil. During the boil flavour changes <strong>and</strong> a<br />
darkening of the colour occurs. Caramels may be added at this stage to adjust the colour.<br />
The hop-boil consumes about half of the energy use in brewing.<br />
1.8 Wort clarification, cooling <strong>and</strong> aeration<br />
At the end of the boil the transparent, or `bright' wort contains flocs of trub (the hot<br />
break) <strong>and</strong> suspended fragments of hops. If whole hops were used then residual solids are<br />
strained off in a hop back or other filtration device <strong>and</strong> the bed of hop cones filters off the<br />
trub, giving a clear, hopped wort. However, if powders, hop pellets, (which break up into<br />
small particles), or extracts were used then hop fragments (if present) <strong>and</strong> the trub are<br />
usually separated in a `whirlpool tank'. The clear `hopped wort' is cooled to check<br />
continuing darkening <strong>and</strong> flavour changes <strong>and</strong> so it can be inoculated (`pitched') with<br />
yeast, <strong>and</strong> can be aerated or oxygenated without a risk of oxidative deterioration. The<br />
heated cooling water is used for various purposes around the brewery. During cooling a<br />
second separation of solids occurs in the wort. This `cold break' is composed mostly of<br />
proteins <strong>and</strong> polyphenols <strong>and</strong> some associated lipids. It is often, but not always,<br />
considered desirable to remove this material to give a `bright', completely clear wort. The<br />
wort is aerated or even oxygenated, to provide oxygen for the yeast in the initial stages of<br />
fermentation.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1.9 Fermentation<br />
Fermentation may be carried out in many different types of vessel (Chapter 14; Boulton<br />
<strong>and</strong> Quain, 2001). Fermenters may be open or completely closed or they may allow part<br />
of the yeast to be exposed to the air for part of the fermentation period. The variety of<br />
fermenters remains because yeasts working in different vessels produce beers with<br />
different flavours. Wort fermentation is initiated by pitching (inoculating) the cooled,<br />
hopped wort with aselected yeast. In afew cases mixtures of yeasts are used. Brewery<br />
yeast is amass of tiny, single, ovoid cells (Saccharomyces cerevisiae, the `sugar fungus<br />
of beer'). Yeast strains vary in their properties <strong>and</strong> the flavours they impart.In avery few<br />
cases, as with Belgian Gueuze <strong>and</strong> Lambic beers, (or some African beers; Chapter 16),<br />
fermentation occurs `spontaneously' <strong>and</strong> a complex mixture of microbes is involved. The<br />
yeast metabolizes extract substances dissolved in the wort. More yeast cells <strong>and</strong> `minor'<br />
amounts of many substances are produced, some of which add to the beer's character.<br />
The major products of carbohydrate metabolism are ethyl alcohol (ethanol), carbon<br />
dioxide <strong>and</strong> heat. The yeast multiplies around 3±5 times. Some is retained for use in<br />
subsequent fermentations, while the surplus is disposed of to distillers or the makers of<br />
yeast extracts.<br />
Traditionally, ales are fermented with `top yeasts' which rise to the top of the beer in<br />
the head of foam. These are pitched at about 16 ëC (61 ëF) <strong>and</strong> fermentation is carried out<br />
at 15 20 ëC (59 68 ëF) for 2±3 days. Traditional lagers are fermented with `bottom<br />
yeasts', which settle to the base of the fermenter. These are pitched at lower temperatures<br />
(e.g., 7 10 ëC; 44.6 50 ëF) <strong>and</strong> fermentations are also carried out at lower temperatures<br />
(e.g., 10 15 ëC; 50 59 ëF), consequently they take longer than ale fermentations. As<br />
wort is converted into beer the removal of materials (especially sugars) from solution <strong>and</strong><br />
the appearance of ethanol both contribute to the decline in specific gravity. The initial or<br />
original gravity, OG, the final or present gravity at the end of the fermentation, FG or PG,<br />
<strong>and</strong> the final alcohol content, are important characteristics of beers.<br />
Yeasts are selected with reference to:<br />
1. their rate <strong>and</strong> extent of growth<br />
2. the rate <strong>and</strong> extent of fermentation<br />
3. the flavour <strong>and</strong> aroma of the beer produced<br />
4. in older fermentation systems it is imperative that top yeasts rise into a good head of<br />
foam <strong>and</strong> bottom yeasts sediment cleanly.<br />
Substances (finings) may be added to promote yeast separation at the end of<br />
fermentation. However, in some modern systems `powdery' yeasts are employed that<br />
stay in suspension until the beer is chilled or until collected by centrifugation.<br />
1.10 The processing of beer<br />
When the main, or `primary' fermentation is nearly complete the yeast density is reduced<br />
to a pre-determined value. The `green' or immature beer (it is not green in colour, but has<br />
an unacceptable, `immature' flavour) is held for a period of maturation or secondary<br />
fermentation. During this process the flavour of the mature beer is refined. Sometimes<br />
`priming' sugar or a small amount of wort is added to boost yeast metabolism <strong>and</strong> the<br />
`maturation', `conditioning' or `lagering' process. (Lagern is German <strong>and</strong> means stored<br />
or deposited). In traditional lager brewing the immature beer was stored cold, e.g., at<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
2 ëC (28.4 ëF), for extended periods, sometimes months, when a very slow secondary<br />
fermentation occurred <strong>and</strong> yeast <strong>and</strong> cold trub settled to the base of the storage vessel.<br />
Conditioning is carried out in various ways. The primary <strong>and</strong> secondary fermentations<br />
were carried out in separate, special vessels but increasingly single vessels are used.<br />
Traditionally, ales are run from fermenters into casks or bottles with a little sugar, finings<br />
<strong>and</strong> a regulated amount of yeast. The secondary fermentation `conditions' the beer in the<br />
container, charging it with carbon dioxide. The ale is dispensed from above a layer of<br />
settled yeast. Such naturally conditioned beers are now made in only small amounts.<br />
These beers are not stable for extended periods <strong>and</strong> they require careful, intelligent<br />
h<strong>and</strong>ling.<br />
Now, after conditioning in bulk, most beers are chilled <strong>and</strong> filtered or centrifuged to<br />
remove residual yeast. These completely bright beers are then carbonated, that is, their<br />
carbon dioxide content is adjusted, they are transferred into bottles, cans, kegs, or bulk<br />
tanks. Nitrogen gas is sometimes added to the package, so the beer contains both this <strong>and</strong><br />
carbon dioxide, but as far as possible air is excluded. Before packaging the beer may be<br />
sterile filtered, a process that avoids flavour damage but it follows that all subsequent<br />
beer movements must be made under rigidly aseptic conditions. More often the beer is<br />
pasteurized, that is, it is subjected to a carefully regulated heat treatment. This may be<br />
applied to the filled bottles or cans or to the flowing beer as it moves to fill a sterile<br />
container. With the notable exceptions of some dark stouts <strong>and</strong> wheat beers, such beers<br />
should (a) be brilliantly clear, (b) develop a stable white foam, or head, when poured into<br />
a clean glass, <strong>and</strong> (c) their flavours <strong>and</strong> gas-contents should remain steady.<br />
The careful selection of raw materials <strong>and</strong> processing conditions help brewers to<br />
approach these objectives. However, it may be necessary to employ other techniques. For<br />
example, the plant proteolytic enzyme papain may be added to beer, or the beers may be<br />
treated with insoluble adsorbents to remove haze precursors. In addition substances may<br />
be added to reduce the dissolved oxygen content of the beer, to maximize its haze <strong>and</strong><br />
flavour stability. Other substances may be added to stabilize beer foam.<br />
1.11 Types of beer<br />
There is no truly satisfactory classification of beers. `Clear, European-style beers' may be<br />
distinguished by the raw materials used in their preparation, the ways in which the<br />
brewing operations are carried out, whether top, bottom or `bulk' fermentation is used,<br />
how the product is conditioned, whether it is chilled <strong>and</strong> filtered <strong>and</strong> carbonated or is<br />
conditioned in bottle or cask <strong>and</strong> how it is packaged. Stouts, porters <strong>and</strong> wheat beers,<br />
which are produced in conventional ways, are often not transparent. A beer may also be<br />
distinguished by its OG <strong>and</strong> degree of attenuation or alcohol content, colour, acidity,<br />
flavour <strong>and</strong> aroma, by its `body' or `mouth feel', by its head (foam) characteristics <strong>and</strong> by<br />
its physiological effects. How a drinker perceives a beer is influenced by many factors,<br />
including the manner in which it is served, its temperature, clarity <strong>and</strong> colour, flavour,<br />
aroma <strong>and</strong> `character', the ambience, <strong>and</strong> whether or not it is being taken with food <strong>and</strong><br />
what has been consumed before.<br />
Within each grouping, `class' or `style', individual beers may be quite distinct <strong>and</strong><br />
brewers aim to produce distinctive products. In North America most beer is pale, lightly<br />
hopped <strong>and</strong> served very cold (often at about 0 ëC; 32 ëF). Many new, small breweries have<br />
been set up <strong>and</strong> these make a wide variety of beers based on styles from around the world.<br />
In Europe, for about a century, British brewing <strong>practice</strong>s diverged from those of mainl<strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
producers, but in recent years convergence has started. For example, in Germany most<br />
beers(notall)weremadeusingdecoctionmashing,bottomfermentation<strong>and</strong>longperiods<br />
of cold storage (lagering). Increasingly, temperature programming, infusion mashing,<br />
bulk fermentation <strong>and</strong> shorter periods of lagering are being used. Under many<br />
circumstances the use of adjuncts is still not used. Although under EEC legislation the<br />
use of adjuncts is allowed, most German brewers still abide by the Reinheitsgebot for<br />
domestic beers.<br />
ThemaingroupsofbeersaretheverypalePilsentypes,palegolden-brownViennatypes,<br />
<strong>and</strong> the darker, rich Munich types. Other beers include MaÈrzen, Oktoberfest, wheat beers,<br />
ryebeers<strong>and</strong>smokedbeers.IntheUKlagerworts areproduced inmanyways,buttheyare<br />
bottomfermented.TheBritish`lagers'areallpalebeers.Alesaretraditionallymadewithan<br />
infusion mashing system. They are moderately strongly hopped <strong>and</strong> atop fermentation<br />
systemisused.Traditionalgroupsarethe(progressivelydarker)paleales,mildales(usually<br />
darker,sweeter<strong>and</strong>lessstronglyhopped),brownales(darkerformsof`mild'),<strong>and</strong>stoutsor<br />
`porters'. The distinctionsbetween ale <strong>and</strong>lager breweries are increasingly blurred assome<br />
brewers adopt similar wort production <strong>and</strong> fermentation systems.<br />
Less common products include wheat beers, low-alcohol <strong>and</strong> alcohol-free beers<br />
(which may be carbonated worts, underfermented beers or beers from which the alcohol<br />
hasbeenremoved),<strong>and</strong>beerswithexceptionallyhighalcoholcontents(e.g.,barleywines<br />
<strong>and</strong> Trappist beers, with 9% ABV, or more). In low carbohydrate (lite, light or dietetic)<br />
beers, prepared by using special mashing conditions <strong>and</strong> added starch-degrading<br />
enzymes, essentially all the starch-derived dextrins are degraded to fermentable sugars<br />
<strong>and</strong> are utilized by the yeast. African opaque beers (Chapter 16) <strong>and</strong> kvass (Russian) are<br />
distinct products. Some unusual beers, made in Belgium, include Lambic, Gueuze <strong>and</strong><br />
fruit-flavoured beers (kriek, flavoured with cherries; framboise, flavoured with<br />
raspberries). These are all made using spontaneous fermentations which involve mixtures<br />
of organisms.<br />
Beer strength may be defined in several ways; by the specific gravity of the wort before<br />
fermentation (the OG), by the alcohol content of the final beer (% alcohol by volume or<br />
ABV) or even by the content of hop bitter substances. The fermentability of extract<br />
depends on many factors. There is no fixed relationship between the OG <strong>and</strong> the alcohol<br />
content of abeer. In Britain the specific gravity of awort or beer is usually quoted times<br />
1,000 so, for example, water has aSG of 1000.00 <strong>and</strong> wort with aspecific gravity (s.g.) of<br />
1.040 at 20ëC (68ëF) has aSG of 1040.00. In the past, extract was calculated as brewer's<br />
pounds per barrel, <strong>and</strong> the excess weight (in lb.) over water was referred to as brewer's<br />
poundsgravity.Thusabarrelofwater(36imp.gallons,UK)weighs360pounds(lb.),buta<br />
barrel of wort at SG 1040 weighed 374.2lb. So this wort had agravity of 14.2lb. Outside<br />
the UK concentrations are often expressed in terms of concentrations of sucrose solutions<br />
of the same gravity (see appendix). Thus, in von Balling's tables of 1843 wort of aspecific<br />
gravity of 1040 is equivalent to a sucrose solution of 9.95% (w/w). Von Balling's tables<br />
were revised by Plato in 1918 <strong>and</strong> gravity is often expressed as degrees Plato. Increasingly<br />
beer strengths are being given as the concentration of alcohol % ABV that they contain.<br />
1.12 Analytical systems<br />
For both trading <strong>and</strong> quality-control purposes all the materials used in making beers, the<br />
liquor, the sweet <strong>and</strong> hopped worts <strong>and</strong> the beers themselves are analysed. Not all the<br />
methods used are st<strong>and</strong>ardized <strong>and</strong>, regrettably, there are at least four `agreed', but<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
discordant, sets of methods in use. The methods used in the different sets differ<br />
significantly <strong>and</strong> give different results. In many instances there are no valid or reliable<br />
conversion factors to interconvert analytical results. The most commonly used methods<br />
are those of the Institute of <strong>Brewing</strong> (IoB; now the Institute <strong>and</strong> Guild of <strong>Brewing</strong>, IGB),<br />
the European Brewery Convention (EBC), the American Society of <strong>Brewing</strong> Chemists<br />
(ASBC) <strong>and</strong> the methods of the MitteleuropaÈischen Analysen Kommission (MEBAK).<br />
The methods are frequently revised, successive versions being distinguished by their<br />
dates.Inthisbookthemostrecent unitsareusedwhereverpossible.By2005themethods<br />
of the IGB <strong>and</strong> the EBC should have been merged. The number of units of measurement<br />
in use is large. Here metric units have been used where possible, with British (UK)<br />
equivalents so, for example, hectolitres <strong>and</strong> imperial gallons. It should be noted that the<br />
American gallon (US) has only about 0.8 of the volume of an imperial gallon. Systems of<br />
units <strong>and</strong> conversion factors are given in the Appendix.<br />
1.13 The economics of brewing<br />
The economics of brewing are influenced by many factors, including the manning levels<br />
required, the local costs of labour, raw materials, how brewing <strong>practice</strong>s are influenced<br />
by governmental regulations <strong>and</strong> how the products are taxed. The scales of brewery<br />
operations vary widely, from units that produce < 10 barrels (imp. brl, approx.16.4<br />
hectolitres, hl) per week to > 30,000 imp. brl (49,092 hl) per week. Thus savings per imp.<br />
brl that are trivial to the small-scale brewer are worthwhile to a larger operator. Breweries<br />
that operate continuously, for 24 hours a day, use their capital investment in plant to the<br />
best effect <strong>and</strong> they can also make other savings, for example, by using heat-recovery<br />
systems that are not suitable for breweries that operate intermittently. There are strong<br />
<strong>and</strong> increasing pressures to minimize water use, to minimize the production of wastes <strong>and</strong><br />
effluents <strong>and</strong> the release of heat <strong>and</strong> odorous gases (such as vapours from hop-boiling),<br />
<strong>and</strong> `greenhouse gases' such as carbon dioxide <strong>and</strong> refrigerants, to utilize raw materials as<br />
efficiently as possible, <strong>and</strong> to utilize fuels <strong>and</strong> power efficiently.<br />
In order to use plant at peak efficiency it is necessary to have it well engineered,<br />
instrumented, automated <strong>and</strong> maintained so that it can operate nearly continuously. To<br />
make such investment worthwhile the capacity of the plant must be large <strong>and</strong>, in<br />
consequence, the manpower needed to produce a given volume of beer is lower than is<br />
needed with less sophisticated plant. The personnel needed to operate modern plant<br />
successfully must be highly trained. Such plant is most efficient at making large volumes<br />
of relatively few beers. Smaller, more labour-intensive plants are often better suited for<br />
making a wide variety of beers in smaller amounts. Large brewing companies tend to<br />
produce fewer beers in larger <strong>and</strong> larger plants. The problems of `product matching', of<br />
trying to make large volumes of one beer in different breweries, are notorious. Smaller<br />
breweries, making smaller volumes, often of more `specialist', <strong>and</strong> even `eccentric'<br />
beers, are appearing all the time. Smaller breweries usually deliver beer over a small area,<br />
<strong>and</strong> so have lower transportation costs relative to larger breweries, which must deliver to<br />
larger areas to market the larger amounts of beer that they produce.<br />
Energy <strong>and</strong> water requirements per unit volume of beer produced vary widely. In part<br />
this is due to differences in the efficiencies of production plants, but it also depends on<br />
the production processes used <strong>and</strong> on how the beer is packaged. Thus decoction mashing<br />
uses more energy than temperature-programmed infusion mashing. Not all breweries<br />
recover heat from the vapours in their mash-cooker or copper-stacks, <strong>and</strong> the efficiency<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
of heat recovery varies with the sophistication of the equipment used. The heat, power<br />
<strong>and</strong> water usage in bottling <strong>and</strong> canning halls (which not all breweries have) is high,<br />
because of the amount of washing carried out, the conveying <strong>and</strong> the heat used by the<br />
pasteurizers.<br />
A widely adopted technique for improving the economics of a brewery is `high gravity'<br />
(HG) brewing, in which concentrated worts are produced <strong>and</strong> processed. The concentrated<br />
beers produced are diluted for sale. Thus a larger volume of beer is produced, per brew,<br />
than would have been the case had the plant been operated in the conventional way. HG<br />
brewing is a technically sophisticated process. There are difficulties with preparing<br />
concentrated worts unless the addition of sugars or syrups to the copper is allowed. Almost<br />
all the stages of the brewing process have to be adjusted, <strong>and</strong> the water used to dilute the<br />
HG beer must be very carefully sterilized, deoxygenated <strong>and</strong> carbonated.<br />
1.14 Excise<br />
Beer is usually taxed. In Britain malt was taxed <strong>and</strong> the regulations imposed, to maximize<br />
the tax receipts, fossilized the malting <strong>and</strong> brewing processes. The malt tax was<br />
withdrawn in 1880, but the styles of beers that had been produced using well modified ale<br />
malts were established as `traditional' <strong>and</strong> continued in use. Only recently have newer<br />
methods of brewing been widely adopted. Next, tax was levied on the gravity <strong>and</strong> volume<br />
of the brewer's wort, after boiling <strong>and</strong> cooling. The consequent economic need to convert<br />
as high a proportion of this wort as possible into saleable beer influenced the designs of<br />
fermentation vessels <strong>and</strong> yeast propagators, the recovery of beer from harvested yeast,<br />
from filters, <strong>and</strong> so on. At present in the UK, <strong>and</strong> many other countries, excise is levied<br />
on the volume <strong>and</strong> alcohol content (ABV) of the beer leaving the brewery. Sometimes<br />
beers are classified according to the alcohol b<strong>and</strong> (range of strengths) in which it falls.<br />
Each b<strong>and</strong> is taxed at a different rate <strong>and</strong> the tax increases with the alcohol content. There<br />
are countries where the beer is taxed by volume only.<br />
1.15 References <strong>and</strong> further reading<br />
1.15.1 The systems of malting <strong>and</strong> brewing analysis<br />
ASBC (1992) The American Society of <strong>Brewing</strong> Chemists. Methods of analysis (8th edn, revised), ASBC,<br />
St. Paul, Minn.<br />
EBC (1997; 1998) European Brewery Convention. Analytica-EBC (5th edn, with revisions), Fachverlag<br />
Hans Carl, NuÈrnberg.<br />
IoB (1997) The Institute of <strong>Brewing</strong>, Recommended Methods of Analysis (2 volumes, <strong>and</strong> revisions), The<br />
Institute <strong>and</strong> Guild of <strong>Brewing</strong>, London.<br />
MEBAK (1993) Brautechnische Analysenmethoden: Methodensammlung der MitteleuropaÈischen<br />
Brautechnischer Analysenkommission (5 volumes).<br />
1.15.2 General<br />
BAMFORTH, C. W. (1998) Beer; tap into the art <strong>and</strong> <strong>science</strong> of brewing, London, Insight Press, 245 pp.<br />
BAMFORTH, C. W. (2002) St<strong>and</strong>ards of <strong>Brewing</strong>: a practical approach to consistency <strong>and</strong> excellence,<br />
Boulder, Colorado, Brewers' Publications, 209 pp.<br />
BOULTON, C. <strong>and</strong> QUAIN, D. (2001) <strong>Brewing</strong> Yeast <strong>and</strong> Fermentation, London, Blackwell Science, 644 pp.<br />
BRIGGS, D. E. (1998) Malts <strong>and</strong> Malting, London, Blackie Academic <strong>and</strong> Professional/Gaithersburg,<br />
Aspen Publishing, 796 pp.<br />
COULTATE, T. P. (2002) Food, the chemistry of its components, (4th edn), Cambridge, The Royal Society<br />
of Chemistry, 432 pp.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HLATKY, C. <strong>and</strong> HLATKY, M. (1997) Bierbrauen zu Hause, Graz, Leopold Stocker Verlag, 177 pp.<br />
HORNSEY, I. S. (1999) <strong>Brewing</strong>, Cambridge, The Royal Society of Chemistry, 231 pp.<br />
HORNSEY, I. S. (2003) A History of Beer <strong>and</strong> <strong>Brewing</strong>, Cambridge, The Royal Society of Chemistry, 742<br />
pp.<br />
KUNZE, W. (1996) Technology, <strong>Brewing</strong> <strong>and</strong> Malting (International edn, translated Wainwright, T.),<br />
Berlin, VLB, 726 pp.<br />
LEWIS, M. L. <strong>and</strong> YOUNG, T. W. (2003) <strong>Brewing</strong> (2nd edn), New York, Kluwer Academic, 398 pp.<br />
MEISEL, D. (1997) A Practical Guide to Good Lager <strong>Brewing</strong> Practice, Hout Bay, South Africa, The<br />
Institute of <strong>Brewing</strong>, Central <strong>and</strong> Southern African Section.<br />
MOLL, M. (1994) Beers <strong>and</strong> Coolers (English edn, translated Wainwright, T.), Andover, Intercept, 495 pp.<br />
SYSILAÈ , I. (1997) Small-Scale <strong>Brewing</strong>. Brew your own beer, Helsinki, Limes, 278 pp.<br />
WAINWRIGHT, T. (1998) Basic <strong>Brewing</strong> Science, Reigate, Wainwright, 317 pp + appendices.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
2<br />
Malts, adjuncts <strong>and</strong> supplementary enzymes<br />
2.1 Grists <strong>and</strong> other sources of extract<br />
The sources of extract used in brewing are materials used in the mash <strong>and</strong> materials<br />
dissolved during the hop-boil (Chapter 1). In addition, small amounts of sugars may be<br />
added to beers as primings or for sweetening. Caramels, coloured malt extracts <strong>and</strong><br />
Farbebier may also be added to adjust colours. Supplementary enzymes, derived from<br />
non-malt sources, may be added to the mash or at later stages of beer production. Malt is<br />
thetraditional source ofenzymes<strong>and</strong>theextractproducedinmashing(Chapters1<strong>and</strong>4).<br />
The contents of this chapter are discussed in more detail elsewhere (Briggs, 1998;<br />
Brissart et al., 2000).<br />
2.2 Malting<br />
2.2.1 Malting in outline<br />
Barley (Hordeum vulgare) is the cereal grain most often malted. Wheat (Triticum<br />
aestivum)<strong>and</strong>sorghum(Sorghumvulgare)arealsomaltedinnotablequantities (thelatter<br />
in Africa), but small amounts of rye (Secale cereale), oats (Avena sativum) <strong>and</strong> millets<br />
(various spp.) are also used. The barley grain or corn has acomplex structure (Briggs,<br />
1978, 1998, Figs 2.1 <strong>and</strong> 2.2), <strong>and</strong> is asingle-seeded fruit (a caryopsis). Barley varieties<br />
differ in their suitabilities for malting. Barley plants are annual grasses. Some are planted<br />
in the autumn (winter barleys) while others are planted in the spring (spring barleys).<br />
Grains are arranged in rows, borne on the head, or ear. The number of rows varies, being<br />
two in two-rowed varieties <strong>and</strong> six in six-rowed forms. In mainl<strong>and</strong> Europe winter<br />
barleys are usually of poor malting quality, but some of the two-rowed winter varieties<br />
grown in the UK (such as Maris Otter, Halcyon <strong>and</strong> Pearl) are of outst<strong>and</strong>ingly good<br />
quality. Good spring malting barleys include Alexis, Chariot, Optic <strong>and</strong> Prisma. Grains<br />
vary in size, shape <strong>and</strong> chemical composition.<br />
It is important to underst<strong>and</strong> that malts consist of mixtures of grains with differing<br />
properties. This heterogeneity, which is reflected in the malt, can give rise to problems in<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Awn<br />
Starchy endosperm<br />
Distal<br />
end<br />
Dorsal side<br />
Aleurone<br />
layer Lemma Pericarp Testa<br />
Sub-aleurone<br />
region Palea<br />
Crushed<br />
cell layer<br />
Ventral furrow side<br />
Acrospire<br />
Embryo<br />
Scutellum<br />
Rootlets<br />
Coleorhiza<br />
Rachilla<br />
Micropylar region<br />
Broken pedicel<br />
Scutellar epithelium<br />
Proximal<br />
end<br />
Fig. 2.1 A schematic longitudinal section of a barley grain, to one side of the ventral furrow <strong>and</strong><br />
the sheaf cells (after Briggs et al., 1981).<br />
Sheaf cells<br />
Pigment<br />
str<strong>and</strong><br />
Dorsal side<br />
Rachilla Ventral<br />
furrow<br />
Ventral side<br />
Lemma<br />
Vascular<br />
bundles<br />
Pericarp <strong>and</strong> testa<br />
Palea<br />
Aleurone layer<br />
Sub-aleurone<br />
layer<br />
Central region<br />
Vascular bundle<br />
Starchy<br />
endosperm<br />
Fig. 2.2 A diagram of a transverse section of a plump barley grain, taken at the widest part (after<br />
Briggs et al., 1981).<br />
brewing. Barley dimensions vary, usually in the ranges: lengths, 6 12 mm, 0.24<br />
0.47 in.; widths, 2.7 5.0 mm, 0.11 0.20 in.; thicknesses, 1.8 4.5 mm, 0.07 0.18 in.<br />
Two-rowed malting barley grains may have one thous<strong>and</strong> corn dry weights (TCW) in the<br />
range 32 44 g, <strong>and</strong> some six-rowed barleys have values of about 30 g. Differences<br />
between grain sizes must be allowed for when setting brewer's mills. The barley corn is<br />
elongated <strong>and</strong> tapers at the ends (Figs 2.1, 2.2). The dorsal, or rounded side is covered by<br />
the lemma, while the ventral, grooved or furrow side is covered by the palea. Together<br />
these units constitute the husk. The lemma has five longitudinal ridges, or `veins' running<br />
along it while the palea has two. In threshed grain the apical tip of the lemma is crudely<br />
broken off. In the unthreshed grain this is where the extended awn is attached. At the base<br />
of the grain, where it was attached to the plant, the rachilla, or basal bristle, lies in the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ventral furrow. Rachillae vary greatly in their shapes <strong>and</strong> sizes, <strong>and</strong> are of use in helping<br />
to identify grain variety. The husk protects the grain from physical damage. In wheat, rye,<br />
sorghum <strong>and</strong> millets (<strong>and</strong> in some few `naked' barleys, which are not malted) husks are<br />
absent in threshed grain, so the corns are easily damaged.<br />
Within the husk the multi-layered pericarp also has a protective function. Finally, the<br />
testa is the layer that `seals' the interior of the grain from the exterior <strong>and</strong> limits the<br />
inward <strong>and</strong> outward movements of dissolved substances, such as sugars, amino acids,<br />
salts <strong>and</strong> proteins. This layer invests the entire interior of the grain except at the embryo,<br />
where its structure is modified in the micropylar region, <strong>and</strong> in the furrow, where the two<br />
edges are sealed together by the pigment str<strong>and</strong>. The testa consists of two cuticularized<br />
layers between which polyphenolic proanthocyanidins usually occur. At the base of the<br />
grain, over the embryo <strong>and</strong> between the pericarp <strong>and</strong> the husk, there are two small, hairy<br />
structures, the lodicules. During steeping these may distribute water over the embryo, by<br />
capillarity. Their varied forms make them valuable aids in identifying a grain's variety.<br />
Within the testa, at the base of the grain, is the small embryo. This is situated towards<br />
the dorsal side of the grain. The embryonic axis consists of the coleoptile (the maltster's<br />
`acrospire') pointing towards the apex of the grain <strong>and</strong> the root sheath (coleorhiza) which<br />
surrounds several (typically five) embryonic roots. This appears at the end of the grain, at<br />
the onset of germination, as the `chit'. The axis is the part of the embryo that can grow<br />
into a small plant. It is recessed into an exp<strong>and</strong>ed part of the embryo called the scutellum<br />
(Latin, `little shield'). Unlike the scutellum in oats, in barley this organ does not grow. Its<br />
inner surface, which is faced with a specialized epithelial layer, is pressed against the<br />
largest tissue of the grain, the starchy endosperm. With the exception of the embryo all<br />
the tissues mentioned so far are dead. All the surface structures, outside the testa, are<br />
infested with mixed populations of micro-organisms.<br />
The starchy endosperm is a dead tissue of thin-walled cells packed with starch<br />
granules embedded in a protein matrix. The granules occur in two size ranges (usually<br />
with diameters 1.7 2.5 m <strong>and</strong> 22.5 47.5 m), which behave differently during malting<br />
<strong>and</strong> brewing. The cell walls are mainly -glucans, with some pentosans <strong>and</strong> a little<br />
holocellulose. This tissue contains most of the grain's reserves, although others are<br />
present in the embryo <strong>and</strong> in the aleurone layer. In transverse section the cell walls radiate<br />
outwards from a `crest' of sheaf cells that run along the grain, above the pigment str<strong>and</strong>.<br />
These sheaf cells are devoid of contents <strong>and</strong> consist of cell walls pressed together, at least<br />
in the dry grain. They are not part of the endosperm tissue, the cell walls of which are<br />
more readily degraded by enzymes (Briggs, 2002). The outer region of the starchy<br />
endosperm, the sub-aleurone layer, is relatively richer in protein (including -amylase)<br />
<strong>and</strong> small starch granules but poor in large starch granules. Where the starchy endosperm<br />
fits against the scutellum the cells are devoid of contents <strong>and</strong> the cell walls are pressed<br />
together, comprising the crushed-cell or depleted layer. The starchy endosperm, away<br />
from the sheaf cells, is surrounded by the aleurone layer (which botanically is also<br />
endosperm tissue). On average it is about three cells thick. The cells are alive but do not<br />
multiply or grow during germination, have thick cell walls <strong>and</strong> contain reserves of lipids<br />
(fat) <strong>and</strong> protein, sucrose <strong>and</strong> possibly fructosans, as well as a full range of functional<br />
organelles. They do not contain any starch. A reduced layer of aleurone tissue, a single<br />
layer of flattened cells, extends partly over the surface of the embryo. The estimates are<br />
approximate, but on a dry weight basis (d.b.) a two-rowed barley corn may consist of<br />
husk + pericarp + lodicules, 9 14%; testa, 1 3%; embryo, 2 3.5%; aleurone layer,<br />
about 5%; starchy endosperm + sheaf cells, 76 82%. Malting can be understood only by<br />
reference to the grain structure <strong>and</strong> the interactions which occur between the tissues.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Barley is purchased in large amounts. The grain delivered must be of the correct<br />
quality, i.e., it must match or exceed in quality asample seen in advance or an agreed<br />
specification. The evaluation of the grain involves both visual <strong>and</strong> laboratory<br />
assessments. Each delivery should be checked before it is unloaded. Delivery may be<br />
by railway, barge or (most usually in the UK) by lorry. The grain will be uncovered <strong>and</strong><br />
inspected for infesting insects, local wetting, admixture of varieties, the presence of ergot<br />
sclerotia (poisonous, grain-sized structures produced by the fungus Claviceps purpurea),<br />
or any sign of heavy fungal attack. If any of these faults is noted the load is likely to be<br />
rejected <strong>and</strong>, if insects are present, the load will be ordered off the premises. With the<br />
exceptionofvarietieswithblue-pigmentedaleuronelayers,(whichappear greenishasthe<br />
blue is viewed through the yellow husk), grain should appear `bright', with aclean strawyellow<br />
colour. Discoloration is caused by heavy microbial contamination.<br />
Samples of the grain bulk are drawn <strong>and</strong> sent to the laboratory. The moisture content<br />
willbedetermined.IntheUKthegrainwillbeinspectedtocheckthatitispredominantly<br />
(e.g., >97%) of one specified variety, that its viability or germinative capacity (GC;<br />
checked by tetrazolium staining) is equal to or exceeds the specified limit (at least 98%)<br />
<strong>and</strong> that the total nitrogen content (TN) or crude protein content (6.25 TN) is within<br />
specified limits. Grain moisture <strong>and</strong> nitrogen contents are usually checked using nearinfra-red<br />
spectroscopy (NIR), but slower methods may be used. The grain will also be<br />
checked for `pre-germination', since grain that has already started to germinate will not<br />
keep or malt well. Asample will be graded (screened) by shaking on aset of slotted<br />
sieves, usually with slot widths of 2.2 or 2.25, 2.5, <strong>and</strong> 2.8mm. In North America the slot<br />
sizes are 7/64 in., 6/64 in. <strong>and</strong> 5/64 in. (about 2.78, 2.38 <strong>and</strong> 1.98 mm, respectively). The<br />
sample must have an acceptable size distribution. Grain, dust <strong>and</strong> rubbish passing the<br />
2.2mm or other agreed screen is regarded as `screenings', or thin corns. It will not be<br />
malted <strong>and</strong> so will have to be removed, collected <strong>and</strong> sold as animal feed. If screenings<br />
exceed aspecified weight percentage the load may be rejected or purchased at areduced<br />
price.<br />
Each lorry-load of grain (typically 20 25t) will be evaluated on afew hundred grams<br />
of grain. For the results to have any statistical validity, because of the inherent<br />
inhomogeneity of grain, the samples must be drawn, mixed <strong>and</strong> sub-divided strictly in<br />
accordancewiththerulessetoutinthesetsofanalyticalmethods(Section1.15.1,p.9).If<br />
the load is acceptable it will be unloaded <strong>and</strong> transferred to a `green grain' store. The<br />
grain is not green in appearance but at this stage it has not been pre-cleaned, dried,<br />
screened or further graded. Grain is best h<strong>and</strong>led <strong>and</strong> stored in batches, separated by<br />
variety, TN, <strong>and</strong> grade. After thorough cleaning, drying <strong>and</strong> perhaps more extended<br />
storage the grain will receive a more thorough laboratory evaluation. These check<br />
procedures take days, compared for the checks carried out at grain intake, for which only<br />
a few minutes are available. Efforts are made to ensure that the grain does not carry<br />
unacceptable levels of residues of insecticides, fungicides, plant growth regulators, or<br />
herbicides by checking the grain's history with suppliers. Some grain samples will be sent<br />
to specialized laboratories to check residue levels.<br />
2.2.2 Changes occurring in malting grain<br />
Before malting, grain is screened <strong>and</strong> aspirated to remove large <strong>and</strong> small impurities <strong>and</strong><br />
`thin' corns. To initiate malting it is hydrated. This is achieved by `steeping', immersing<br />
the grain in water or `steep liquor'. Later, the moisture content may be increased by<br />
spraying or `sprinkling' the grain. The steep-water temperature should be controlled. At<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
elevated temperatures water uptake is faster but microbial growth is accelerated <strong>and</strong> the<br />
grain may be damaged or killed. The best temperature for steeping immature (partly<br />
dormant) grain is low (about 12 ëC, 53.6 ëF). For less dormant grain a value of 16 18 ëC<br />
(60.8 64.4 ëF) is often used. As the grain hydrates it swells to 1.3 1.4 times its original<br />
volume. To prevent it packing tightly <strong>and</strong> wedging in the steep it may be loosened <strong>and</strong><br />
mixed by blowing air into the base of the steeping vessel. This also adds oxygen to the<br />
steep liquor. The oxygen is rapidly taken up, both by the grain <strong>and</strong> by the microbes that<br />
multiply on the grain <strong>and</strong> in the liquor. Material is leached from the grain <strong>and</strong> enzymes<br />
from the microbes start to degrade the materials in the grain surface layers. Thus the<br />
liquor contains an increasing number of microbes, microbial metabolites <strong>and</strong> dissolved<br />
substances, it becomes yellow, gains a characteristic smell <strong>and</strong> may froth.<br />
Some of the substances <strong>and</strong> the microbes in the liquor check grain germination.<br />
Infestations of microbes are undesirable. They compete with the grain for oxygen <strong>and</strong><br />
reduce the percentage germination <strong>and</strong> germination vigour. Some produce plant growth<br />
regulators (including gibberellins) which stimulate or inhibit malting, others may produce<br />
mycotoxins which damage yeasts <strong>and</strong>/or are toxic to human beings. Some produce agents<br />
which cause beer to gush (over-foam), they produce some hydrolytic enzymes which may<br />
improve malt performance in the mash tun. Contamination with bacteria may give rise to<br />
worts <strong>and</strong> beers which are hazy with suspended, dead microbes (Schwarz et al., 2002;<br />
Walker et al., 1997).<br />
Steep water, which checks grain germination <strong>and</strong> growth if re-used, is periodically<br />
drained from the grain <strong>and</strong> replaced with fresh. The minimum acceptable number of<br />
water changes are used since both the supply of fresh water <strong>and</strong> the disposal of steep<br />
effluent are costly. Sterilants are not routinely used in steeps, but many substances,<br />
including mineral acids, potassium <strong>and</strong> sodium hydroxides, potassium permanganate,<br />
sodium metabisulphite, slaked lime water <strong>and</strong> slurried calcium carbonate <strong>and</strong><br />
formaldehyde, have been used, as has hydrogen peroxide. `Plug rinsing' grain in the<br />
steep by washing downwards with a layer of fresh water, (with or without hydrogen<br />
peroxide or other substances), as the steep is drained is an economical possibility for<br />
removing suspended microbes, their nutrients <strong>and</strong> other substances (Briggs, 2002).<br />
To control the microbes which produce mycotoxins <strong>and</strong> gushing-promoting agents it<br />
has been proposed that they should be swamped with `harmless' microbes which will<br />
outgrow the problem-causing species. Species investigated include lactobacilli <strong>and</strong><br />
strains of Geotrichum yeast (Boivin <strong>and</strong> Mal<strong>and</strong>a,1998; Haikara et al., 1993; Laitila et<br />
al., 1999). The results appear promising, but these microbes will also compete with the<br />
grain for oxygen. Their use might be combined with a washing procedure (Briggs, 2002).<br />
Air rests are used between steeps. After a steep has been drained air, which should be<br />
humid <strong>and</strong> at the correct temperature, is sucked down through the grain. Such downward<br />
ventilation, or `CO2 extraction', assists drainage, provides the grain with oxygen,<br />
removes the growth-inhibiting carbon dioxide <strong>and</strong> removes some of the heat generated by<br />
the metabolizing grain. In consequence, <strong>and</strong> in contrast to traditional <strong>practice</strong>, barley<br />
leaving the steep has usually started to germinate. When the grain is immersed it is partly<br />
anaerobic, <strong>and</strong> it ferments, forming carbon dioxide <strong>and</strong> alcohol (ethanol), a proportion of<br />
which enters the steep liquor. Under such conditions the grain will not germinate. Under<br />
aerobic conditions fermentation is repressed <strong>and</strong> germination can occur. During<br />
immersions air may be blown into the base of a steep, providing some oxygen <strong>and</strong><br />
lifting <strong>and</strong> mixing the grain.<br />
The onset of germination is indicated by the appearance of the small, white `chit', the<br />
root sheath (coleorhiza) that protrudes from the base of each germinated grain. At this<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
stage the grain is transferred to a germination vessel (or floor in older maltings) or, if it is<br />
in a steeping/germination vessel, the equipment will be set into the germination mode.<br />
The grain grows, producing a tuft of rootlets (culms) at the base of the grain <strong>and</strong>, less<br />
obviously, the coleoptile or `acrospire' grows along the dorsal side of the grain, beneath<br />
the husk. The extent of acrospire growth, expressed as a proportion of the length of the<br />
grain, is used as an approximate guide to the advance of the malting process. Variations<br />
in acrospire lengths indicate heterogeneity in growth. The living tissues respire <strong>and</strong><br />
carbon dioxide <strong>and</strong> water are generated resulting in a loss of dry matter. The energy<br />
liberated supports growth <strong>and</strong> is liberated as heat.<br />
Many hydrolytic enzymes, which are needed when malt is mashed, appear or increase in<br />
amount. Some of these catalyse the physical modification of the starchy endosperm. In the<br />
initial stages of germination these hydrolases are released from the scutellum. However,<br />
after a short lag the embryo releases gibberellin hormones (GA 1 <strong>and</strong> GA 3, gibberellic acid).<br />
These diffuse along the grain triggering the formation of some enzymes in the aleurone layer<br />
<strong>and</strong> the release of these <strong>and</strong> other enzymes into the starchy endosperm. Here they join the<br />
enzymes from the embryo in catalysing modification. As germination progresses the starchy<br />
endosperm softens <strong>and</strong> becomes more easily `rubbed out' between finger <strong>and</strong> thumb. When<br />
the malt has been dried the modified material is easily crushed <strong>and</strong> `friable', <strong>and</strong> is easily<br />
roller-milled, in contrast to the tough barley. The stages of physical modification are the<br />
progressive degradation of the cell walls of the starchy endosperm, which involves the<br />
breakdown of the troublesome -glucans <strong>and</strong> pentosans, followed by the partial degradation<br />
of the protein within the cells <strong>and</strong> the partial or locally complete breakdown of some of the<br />
starch granules, the small granules being attacked preferentially. The extent of breakdown is<br />
limited by the availability of water.<br />
Modification begins beneath the entire `face' of the scutellum. In a proportion of<br />
grains it advances more rapidly on the ventral side of the endosperm, adjacent to the sheaf<br />
cells, while in others it advances roughly parallel to the face of the scutellum (Briggs,<br />
1998). When enzymes from the aleurone layer have been produced modification<br />
progresses more rapidly, particularly adjacent to the aleurone layer. Thus modification<br />
begins adjacent to the embryo <strong>and</strong> advances towards the apex as germination proceeds. In<br />
well-made malt only a small proportion of grains are undermodified, <strong>and</strong> contain large<br />
amounts of undegraded (unmodified) starchy endosperm tissue. The products of<br />
endosperm breakdown, sugars, amino acids, etc., together with materials from the<br />
aleurone layer (phosphate, metal ions, etc.), diffuse through the endosperm <strong>and</strong> a<br />
proportion support the metabolism of the living tissues, while the remainder accumulates.<br />
The growth of the embryo is at first supported by its own reserve substances <strong>and</strong> later<br />
by soluble materials from the modifying starchy endosperm, so there is a net migration of<br />
materials into the embryo. The levels of soluble materials that accumulate are regulated<br />
by the balance between their rates of formation in the endosperm <strong>and</strong> their rates of<br />
utilization by the embryo. In the finished malt these materials may be estimated as the<br />
cold water extract, CWE, or the `pre-formed solubles'. The accumulation, with time, of<br />
enzymes <strong>and</strong> the physical modification of the grain, permit the increasingly greater<br />
recovery of hot water extract up to a maximum value. When the acrospires have grown to<br />
about 3/4 to 7/8 the length of the grain the hot water extract, the cold water extract <strong>and</strong><br />
the level of soluble nitrogenous substances cease to increase with increasing germination<br />
time, <strong>and</strong> the fine-coarse extract difference has almost stopped decreasing although<br />
friability is still increasing <strong>and</strong> the viscosity of grain extracts may still be declining.<br />
Enzyme levels may or may not be increasing, depending on the malting conditions.<br />
Usually germination is terminated at this stage by kilning. Longer germination periods<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
waste malthouse capacity <strong>and</strong> result in extra malting losses. A correct dose of gibberellic<br />
acid, GA3, which, when used, is usually applied at the end of steeping, accelerates the<br />
growth of the embryo but stimulates most of the processes of modification relatively<br />
more so that malt can be prepared more rapidly <strong>and</strong> in better yield. Where permitted the<br />
excessive accumulation of soluble nitrogenous substances that can occur in GA 3-treated<br />
grain may be limited by the application of sodium or potassium bromate, which inhibits<br />
the activity of some proteolytic enzymes. By checking the growth of rootlets this agent<br />
also increases malt yields.<br />
The processes that occur in germination are regulated by controlling the moisture<br />
content of the grain, the quality of the grain <strong>and</strong> the temperature programme of the grain<br />
during steeping <strong>and</strong> the germination period. Nearly all malt is made using `pneumatic'<br />
malting plant in which the grain is ventilated with a stream of humidified <strong>and</strong><br />
temperature-adjusted air to remove excess heat <strong>and</strong> carbon dioxide <strong>and</strong> to supply oxygen.<br />
In the UK, limited amounts of high-quality malts are still made by traditional floor<br />
malting. From time to time the piece (batch) will be turned or stirred to separate the<br />
grains by untangling the rootlets <strong>and</strong> allowing the easier passage of the conditioning<br />
airflow.<br />
Malting losses can be defined in several ways. If they are defined in terms of the losses<br />
in dry weight, which occur when cleaned barley entering the steep is recovered as kilned<br />
malt <strong>and</strong> has been de-culmed (dressed), then the losses sustained in making conventional<br />
malts are usually in the ranges: steeping losses, 0.5 1.5%; germination losses,<br />
3.5 7.5%; rootlets, 2.5 5.0%. These divisions are artificial, since some respiration<br />
<strong>and</strong> growth occur in the steeping phase <strong>and</strong> in the initial stages of kilning. Rootlets are<br />
sold, usually for use in animal feeds, but the cash value is less than that of an equal<br />
weight of malt. Malting losses are larger when coloured malts are being produced.<br />
Modification of the green malt is likely to be more extensive if it is to be used in making<br />
darker malts. It will probably have been germinated with a relatively high moisture<br />
content <strong>and</strong> will be rich in soluble sugars <strong>and</strong> soluble nitrogenous compounds that will<br />
react during kilning to generate melanoidins, <strong>and</strong> so generates colours <strong>and</strong> characteristic<br />
flavours <strong>and</strong> aromas.<br />
Most modern kilns hold a bed of grain about one metre deep, (which is not turned<br />
during kilning), through which a current of air is fan-driven from below. This air is heated<br />
either directly, using low-NOx burners fuelled with oil or gas (which generate little or no<br />
oxides of nitrogen), or indirectly by heat exchangers. Oxides of nitrogen are avoided to<br />
prevent the formation of potentially harmful nitrosamines. When making pale malts the<br />
airflow is rapid <strong>and</strong> the `air-on' temperature is low during the initial, drying phase. As the<br />
air rises through the bed of malt it becomes saturated with moisture <strong>and</strong> it is cooled by the<br />
need to provide energy to evaporate the water. So above the drying zone the air is<br />
saturated with moisture <strong>and</strong> the grain can continue to grow, generate enzymes <strong>and</strong> modify<br />
while being warm. In the initial stages the air-on temperature may be about 50 ëC<br />
(122 ëF), <strong>and</strong> the air-off temperature is around 25 ëC (77 ëF), which will be the<br />
temperature of all the green malt above the top of the drying zone. To economize with<br />
fuel this air should be passed through a heat exchanger to pre-heat incoming air. As the<br />
outgoing air is cooled moisture condenses on the tubes of the heat exchanger, liberating<br />
its heat of condensation, which is passed to the incoming air.<br />
With the passage of time the drying zone extends upwards through the bed of malt<br />
until it reaches the surface of the grain bed. At this time, at the `break point', the relative<br />
humidity of the `air-off' falls <strong>and</strong> the temperature rises. When this occurs the airflow is<br />
reduced <strong>and</strong> the air-on temperature is increased to begin the curing (cooking) stage. As<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the malt dries the temperature is progressively increased to the maximum, `curing'<br />
temperature. As the fuel used in kilning is costly, the procedure is adjusted to save heat.<br />
During curing a progressively higher proportion of the air may be re-circulated.<br />
Alternatively the hot air may be diverted to asecond, `linked' kiln in which the malt is in<br />
the drying stage. Here it is mixed with more heated air to provide the large volume of air<br />
required during drying. Many pale malts are cured at about 80ëC (176ëF), but some will<br />
be `finished' at higher temperatures, up to 105ëC (221ëF).<br />
While enzyme destruction occurs at these elevated temperatures some enzymes<br />
survive provided that the malt has first been dried at low temperatures to alow moisture<br />
content. Under these conditions colour formation is minimized. In the manufacture of<br />
some coloured malts the temperature is increased while the grain is still comparatively<br />
wet to promote the formation of free sugars <strong>and</strong> amino acids <strong>and</strong> the interaction of these<br />
<strong>and</strong> other substances form the coloured melanoidins <strong>and</strong> flavoursome <strong>and</strong> aromatic<br />
substances. In these malts enzyme levels are comparatively low <strong>and</strong>, in extreme cases,<br />
enzyme destruction is complete.<br />
Most special malts are now `finished' in roasting drums. These metal cylinders may<br />
have capacities ranging from 0.5tto 10t. As they turn the contents are mixed by internal<br />
vanes <strong>and</strong> may be heated indirectly, by heating the outside of the drum, or directly, when<br />
hot air is passed through the interior of the cylinder, drying the contents. Depending on<br />
the malt being made adrum is loaded either with pale, kilned malt or green malt, <strong>and</strong> the<br />
processing is varied. The cylinder is heated while rotating <strong>and</strong> the contents are subjected<br />
to acarefully chosen temperature regime. The colour <strong>and</strong> physical state of the grain is<br />
frequently checked on samples. At exactly the correct time heating is stopped, in some<br />
instances water is sprayed into the cylinder, <strong>and</strong> the malt is withdrawn <strong>and</strong> cooled. Green<br />
malt is usually used in making crystal or caramel malts (which are not all dark in colour)<br />
while pale, kilned malts are used in making other types which vary in colour from amber<br />
to black (Briggs, 1998 <strong>and</strong> Sections 2.2.5, 2.2.6).<br />
After kilning malts are dressed (de-culmed or de-rooted <strong>and</strong> cleaned). The cooled malt<br />
is agitated to break up the brittle rootlets <strong>and</strong> these, <strong>and</strong> dust, are separated by sieving <strong>and</strong><br />
aspiration with air currents. Pale malts are usually stored for 4 6 weeks before use when,<br />
for unknown reasons, the brewing values often improve. Coloured <strong>and</strong> special malts<br />
should be brewed with as soon as possible, because during storage their special aromas<br />
(<strong>and</strong> perhaps flavours) decline. Malts are stored in ways intended to minimize the pickup<br />
of moisture, <strong>and</strong> to exclude birds, rats, mice <strong>and</strong> insects. It is important to prevent malts<br />
being mixed or being contaminated with un-malted barley during h<strong>and</strong>ling or storage. It<br />
is impossible to make successive batches of malt that have precisely the same analysis.<br />
Each batch should be stored separately <strong>and</strong> different batches should be blended so that the<br />
mixture meets the brewer's requirements. Different batches of malt made from one<br />
variety of barley, in the same way <strong>and</strong> intended to meet the same specification can be<br />
safely blended. Brewers take different views regarding what other blending is<br />
permissible. Specifications may stipulate that no other blending should occur, while<br />
others will accept blends of any varieties (even involving malts from two- <strong>and</strong> six-rowed<br />
barleys) provided that the analyses of the mixture are as stipulated.<br />
Before dispatch the malt will be cleaned by screening, aspiration, passage over<br />
magnetic separators to remove fragments of iron, <strong>and</strong> (often) through a gravity separator/<br />
de-stoner. Usually malt is delivered in bulk, (often 25 t batches), but for export some<br />
batches will be packed in very large sacks (1 t capacity) which will be transported in<br />
containers. Some maltings will provide smaller breweries with malt in smaller sacks that<br />
are made with several layers of different materials <strong>and</strong> are strong <strong>and</strong> waterproof.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Rootlets are usually sold for cattle feed. However, they have been used to provide<br />
nutrients for microbial cultures <strong>and</strong> in making composts for growing mushrooms. Other<br />
markets are being sought. Because of their low bulk density, inconvenience in h<strong>and</strong>ling<br />
<strong>and</strong> a strong tendency to pick up moisture, rootlets are now usually pelletized, together<br />
with malt <strong>and</strong> grain dust <strong>and</strong> sometimes with thin grains. Rootlets (culms, coombes,<br />
cummins, malt sprouts) vary in their nature depending on what malt they came from <strong>and</strong><br />
in particular how strongly they were kilned (Briggs, 1978,). Commonly analyses are in<br />
the ranges: non-protein extract, 35 50%; crude protein, 20 35%; ash, 6 8% <strong>and</strong> fibre,<br />
9 15%. They are rich in low molecular weight nitrogenous substances <strong>and</strong> B vitamins.<br />
2.2.3 Malting technology<br />
Many types of malting plant are in use, but only the most common types will be<br />
described. Malting is comparatively safe, provided that certain precautions are observed.<br />
Some, sometimes unfamiliar, risks are due to carbon dioxide <strong>and</strong> to grain <strong>and</strong> malt dust.<br />
As grain is steeped <strong>and</strong> germinated it liberates carbon dioxide. This heavy gas can `pool',<br />
so it is essential to check that vessels <strong>and</strong> confined spaces are ventilated before they are<br />
entered. Dust must be confined <strong>and</strong> cleaned away not only because it becomes damp <strong>and</strong><br />
a breeding ground for insects <strong>and</strong> microbes, but also because when it is breathed it can<br />
cause allergies <strong>and</strong> fungal lung infections <strong>and</strong> it can form explosive mixtures when mixed<br />
with air. All h<strong>and</strong>ling equipment must be earthed (grounded) to prevent sparks, which<br />
might trigger an explosion, <strong>and</strong> all conveyors, ducts, etc., should have explosion vents.<br />
Modern malt factories process large batches of grain, (often 200 300 t batches), <strong>and</strong><br />
the process stages are highly automated so that processing conditions are reproducible<br />
<strong>and</strong> the manpower needed to produce each tonne of malt is minimized. In large maltings<br />
grain is delivered in bulk, by ship or barge or train or lorry. Before unloading begins the<br />
bulk should be inspected <strong>and</strong> sampled for analysis. When the quality has been agreed<br />
unloading begins. Grain is usually sucked from the holds of vessels, <strong>and</strong> this pneumatic<br />
system may be used to empty rail wagons or lorries, or these may be emptied under<br />
gravity. Lorries usually unload into an intake pit by tipping or from a hopper. The grain<br />
runs into the pit, which is ventilated to remove dust, <strong>and</strong> is equipped with a coarse<br />
moving screen (sieve) to catch <strong>and</strong> remove coarse impurities such as straw <strong>and</strong> large<br />
stones. Each lorry is weighed on to the site <strong>and</strong> off after unloading. The difference in<br />
weights gives the amount of grain unloaded.<br />
During malting grain will be moved several times. The equipment used varies, but will<br />
usually include bucket elevators, helical screw conveyors (worms), belt conveyors <strong>and</strong><br />
chain <strong>and</strong> flight conveyors. Less usually the grain will be moved using pneumatic<br />
conveyors. It is highly desirable that, to avoid cross-contamination, the equipment used to<br />
move grain is entirely separate from that used to move malt. The freshly delivered barley<br />
is conveyed to a `green grain' bin for temporary storage. Here it will remain, usually<br />
being ventilated with fresh air, until it can be precleaned <strong>and</strong>, if necessary, dried.<br />
Precleaning involves rapid screening to remove gross impurities, such as s<strong>and</strong>, straw,<br />
stones <strong>and</strong> string, which are either appreciably larger or smaller than the grains, <strong>and</strong><br />
aspiration with air to remove dust. The dust from this <strong>and</strong> other locations is trapped in<br />
cyclones <strong>and</strong> textile-sleeve filters. The grain also passes over magnetic separators, which<br />
retain iron <strong>and</strong> steel impurities.<br />
In Northern Europe the grain usually needs to be dried (to 12% moisture, or less)<br />
before it can be safely stored. Drying <strong>and</strong> pre-cleaning may be carried out before the<br />
grain is delivered but, because of the risk of heat damage caused by inexpert drying, some<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
maltsters do not allow this. The drying temperatures used are lower for more moist grain,<br />
because wetter grain is more easily damaged by heat. Batch drying can be carried out in<br />
malt kilns or in steeping, germination <strong>and</strong> kilning units (vessels; SGKVs) or in dedicated<br />
batch driers. In these the grain rests on a perforated floor or deck <strong>and</strong> warm air is passed<br />
through it, e.g., for eight hours, until the grain has been dried sufficiently. The grain then<br />
may or may not be cooled, depending whether it is to be committed to long-term storage<br />
or it is to be stored warm for a short period to overcome dormancy (i.e. to hasten postharvest<br />
maturation). In flow-through dryers the grain passes downwards under gravity in<br />
a stream that is regulated by valves. The grain passes through a series of zones in which it<br />
meets air at different temperatures <strong>and</strong> is successively warmed, dried <strong>and</strong> cooled. If there<br />
is to be a period of warm storage the cooling may be limited or omitted, so that the grain<br />
reaching store is at 30 40 ëC (86 104 ëF), rather than 15 ëC (59 ëF) or less, which is<br />
desirable for long-term storage.<br />
The dried grain may now be thoroughly cleaned either immediately or after warm<br />
storage. This process is less rushed than pre-cleaning <strong>and</strong> so is more thorough. The grain<br />
is screened to remove thin corns <strong>and</strong> sometimes it is graded into size classes (e.g., above<br />
<strong>and</strong> below 2.5 mm width), which are malted separately. The screens used may be flat <strong>and</strong><br />
oscillate horizontally or they may be rotating cylinders. At present the quality of the grain<br />
on delivery in the UK is so good that apart from aspiration, screening <strong>and</strong> passage over<br />
magnetic separators, this is all the cleaning required. However, with less clean samples it<br />
may be necessary to remove light impurities with air classification <strong>and</strong> foreign seeds <strong>and</strong><br />
broken grains with Trieur cylinders or Carter-Simon disc separators (Briggs, 1998). The<br />
clean barley may be stored in flat-bed stores, bins or silos. If storage is to be for an<br />
extended period then the grain can be treated with an approved insecticide. If the grain is<br />
held relatively moist (> 12%) it will have to be ventilated. At a 12% moisture content<br />
grain can be stored for some months at or below 15 ëC, but for periods over about six<br />
months a moisture content of 10% is safer. Stores must be regularly inspected for signs of<br />
insect infestation <strong>and</strong> fungal attack <strong>and</strong> depredations by birds or rodents. The temperature<br />
of the grain, determined by probes positioned at various sites <strong>and</strong> depths, should be<br />
recorded weekly <strong>and</strong> any undue increase acted on as a sign of deterioration.<br />
Grain is weighed on its way to the steep(s). If abrasion (limited physical battering or<br />
rubbing the grains together) is to be employed this is carried out in advance of steeping as<br />
grain can be treated at rates of only 10 12 t/h <strong>and</strong> malting batch sizes are often as high as<br />
300 t, <strong>and</strong> so this amount of treated grain must be accumulated before steeping can begin.<br />
Historically, steeps were barrels or shallow troughs in which grain rested, under water, at<br />
depths of 1 2 ft. (0.31 0.62 m). Numerous patterns of steeping vessels have been used.<br />
Those preferred now are either flat-bed or conical-bottomed steeps. Flat-bed steeps are<br />
circular in plan view, <strong>and</strong> the grain is supported on a perforated deck above the true base, so<br />
there is a plenum beneath the deck. For a 200 t batch size the steep might have a diameter of<br />
15 m (49.2 ft.; Gibson, 1989). Initial depths (before the grain swells) may be 1.5 1.8 m<br />
(approx. 4.9 5.9 ft.). Grain is loaded in from above, dropping through sprays of water that<br />
quench the dust, falling into water. The bed is levelled with a rotating spreader, called a<br />
giracleur. Air may be blown in beneath the deck while the grain is immersed, <strong>and</strong> in dry<br />
periods air may be sucked down through the grain. The steeped grain is discharged through<br />
ports impelled by the giracleur. Such steeps allow relatively even grain treatment, since the<br />
bed depth is comparatively shallow, but the water used to fill the plenum is `waste' <strong>and</strong> so<br />
effluent volumes are large. In addition it is difficult to keep the plenum chamber clean.<br />
Newer maltings usually employ various types of conical-bottomed steeps. As each<br />
steep should contain less than 50 t, to avoid deep cones with excessively high pressures<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
on the grain in the cone base, aset of steeps is required for each batch of grain. Typically<br />
each steep consists of avertical cylinder, closed below with acone. Grain is loaded into<br />
each steep from above. The base of the cone contains avalve that retains the grain <strong>and</strong><br />
water, aperforated zone through which the water can be drained while the grain is<br />
retained, an outlet through which air can be drawn during CO 2extraction during air rests<br />
<strong>and</strong> inlets through which compressed air can be supplied when the grain is being aerated<br />
whenunderwater.Suchsteepsare`self-emptying'.Theconeangleissufficientlyacuteto<br />
ensure that when the valve is opened the grain falls out in to aconveyor. This is one way<br />
of `dry-casting' the grain. An alternative method is to `wet-cast' the grain, pumping it to<br />
the germination compartment slurried in water. In steep-germination <strong>and</strong> steepgermination-kilning<br />
vessels (SGVs <strong>and</strong> SGKVs) no transfer is required. If additives,<br />
such as gibberellic acid <strong>and</strong>/or sodium bromate are to be added it is convenient <strong>and</strong><br />
economic to spray on solutions as the grain is conveyed from the steep.<br />
In floor malting the steeped grain is spread on afloor in aroom having acool, humid<br />
atmosphere. Germination is controlled by turning the `piece' (batch) <strong>and</strong> thickening or<br />
thinning the layer of grain to allow temperature rises or falls as needed. Fine malts can be<br />
made in this way, but only in small quantities (ca. 10t/batch) <strong>and</strong> with substantial<br />
manpower. Modern maltings are of the pneumatic type, in which the grain is turned<br />
mechanically <strong>and</strong> the grain temperature is controlled by forcing astream of attemperated<br />
<strong>and</strong> water-saturated air through abed of grain. Newer germination vessels are usually<br />
rectangular `Saladin boxes' or circular compartments. In these vessels steeped grain is<br />
formed into abed, usually 0.6 1.0m(approx. 2.0 3.3ft.) deep. The grain rests on a<br />
perforated deck, through which the conditioning airflow is driven. Some of the air is<br />
recirculated<strong>and</strong>mixedwithfreshair.Theairisdrivenbyafan<strong>and</strong>isusuallyhumidifiedby<br />
passagethroughspraysofwater.Airtemperaturemaybecontrolled,byregulatingthewater<br />
temperature, sometimes augmented with heating or cooling by heat exchangers. The grain<br />
lifted <strong>and</strong> partly mixed, <strong>and</strong> the rootlets are separated by passing arow of vertical, contrarotating<br />
helical screws through the bed. The bed is `lightened' <strong>and</strong> the resistance to the<br />
airflow is reduced. Bed temperatures of 15 19ëC (59 66.2ëF) are common, with<br />
temperaturedifferentialsbetweenthetop<strong>and</strong>bottomofthebedof2 3ëC(3.6 5.4ëF).The<br />
turner arrays are usually fitted with sprays to allow the grain to be moistened.<br />
In some plants the grain is first germinated in acircular, stainless-steel lined vessel,<br />
then it is transferred to a germination <strong>and</strong> kilning unit (or vessel; GKV). When<br />
germination is sufficiently advanced the cool airflow, which may or may not be<br />
humidified, is discontinued <strong>and</strong> hot air is supplied from afurnace or heat exchanger. In<br />
old kilns the malt was dried in thin layers <strong>and</strong> with periodic turning. In modern kilns the<br />
grain beds are relatively deep <strong>and</strong> are not turned. The kilns may be directly or indirectly<br />
heated. They are instrumented so that correct temperature differentials are maintained<br />
between the air-on <strong>and</strong> the air-off <strong>and</strong> that at the break point the airflow is reduced <strong>and</strong><br />
subsequent air re-circulation, temperatures <strong>and</strong> flow rates are correct. As noted before,<br />
heat should be conserved by `linking' kilns <strong>and</strong>/or using heat exchangers. Further heat<br />
can be recovered from the outgoing air by using aheat pump, but at present this is not<br />
economic because the capital <strong>and</strong> maintenance costs are high.<br />
2.2.4 Malt analyses<br />
Malt analyses are carried out according to one of the several sets of agreed methods<br />
(Section 1.15.1, p. 9). As with barley, analysis of amalt lot is useless unless the samples<br />
used are properly drawn <strong>and</strong> h<strong>and</strong>led. Because of differences between the methods, both<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
in the conditions <strong>and</strong> calculations used, the results obtained often differ, both in the<br />
values obtained <strong>and</strong> the ways in which the results are expressed. Some of the most<br />
important methods will be discussed. Others are considered elsewhere (Briggs, 1987;<br />
1998). The moisture contents of malts are usually in the range 1.5 6%, expressed on a<br />
fresh weight (fr. wt.; as is) basis. Primary analyses are by oven-drying methods, but NIR<br />
(near infra-red) analysis is the usual, more rapid, secondary method. Brewers normally<br />
specify an upper moisture limit. Because malt is hygroscopic it will normally have a<br />
lower value when dispatched, to allow for moisture uptake while in transit. Brewers use<br />
malt `as is', <strong>and</strong> so they pay attention to the extract of the undried malt. However, for<br />
comparative purposes extracts are mostly given on a dry weight basis (on dry). A malt<br />
sample must not contain more than a certain percentage of thin corns, because thin corns<br />
are not broken up in mills with rollers set relatively far apart to achieve a coarse grist.<br />
When malt is hammer milled this consideration does not apply. Cold water extracts<br />
(CWE) are used by some ale brewers. The value of this determination is disputed. The<br />
grain is ground <strong>and</strong> extracted at 20 ëC (68 ëF) with water made alkaline with ammonia (to<br />
inactivate enzymes). The specific gravity of the extract is a measure of the `preformed<br />
soluble substances' present in the malt. Departure from customary values warns that the<br />
malt lot is different from its predecessors.<br />
The hot water extract or extract (HWE or E) value is the single most important<br />
measurement in judging malt quality. The HWE method used by the IoB was designed<br />
for use by traditional ale brewers, <strong>and</strong> involves an isothermal laboratory mash (65 ëC;<br />
149 ëF) made with distilled water <strong>and</strong> a comparatively coarsely ground grist. After one<br />
hour the mash is cooled <strong>and</strong> adjusted to either a volume of 515 ml or to a weight of 450 g<br />
<strong>and</strong> the specific gravity of the liquid, obtained by filtration, is measured at 20 ëC (68 ëF).<br />
Using the appropriate formula the extract is calculated from the excess specific gravity<br />
(i.e., the gravity 1000 above water, taken as 1000.00 at 20 ëC (68 ëF)) as litre-degrees/<br />
kg malt (l ë/kg). Sometimes the HWE of a finely ground grist is also determined. Extract<br />
is obtained in greater yields from finely ground malt <strong>and</strong> the smaller the fine-coarse (f.-c.)<br />
extract difference the better the malt is modified. Because this value is the difference<br />
between two large numbers <strong>and</strong> is small relative to the errors involved in measuring the<br />
extracts, the determination must be replicated to obtain a reliable value, which is<br />
laborious. Another `non-st<strong>and</strong>ard' proposal is to use the difference in extract yield from a<br />
fine grind mash <strong>and</strong> a concentrated, very coarse grind mash, This `f.-c. conc.-extract<br />
difference' method gives larger differences than the f.-c. grind method <strong>and</strong> appears to be<br />
an improvement on it, the values obtained being inversely related to extract recoveries in<br />
a brewery (Bourne <strong>and</strong> Wheeler, 1982; Briggs, 1998).<br />
The determination of extract, E, by the EBC <strong>and</strong> the very similar ASBC methods<br />
differs considerably from the IoB method. They were developed for traditional lager<br />
brewers but the temperature programme used does not resemble that of most old lager<br />
breweries or that of breweries which employ temperature programmed mashing. In the<br />
EBC method finely ground malt is mashed in at a low temperature (45 ëC; 113 ëF), with<br />
continuous stirring. The temperature is then increased, at 1 ëC (1.8 ëF)/min., until it<br />
reaches 70 ëC (158 ëF). This temperature is now maintained <strong>and</strong> more water, also at 70 ëC,<br />
is added. After one hour, during which the `saccharification time' is determined (see page<br />
24), the mash is cooled <strong>and</strong> adjusted to 450 g. The specific gravity of the wort is<br />
determined. Using tables that relate the strengths of sucrose solutions with their specific<br />
gravities, the weight of extract in the laboratory wort is calculated, assuming that the<br />
dissolved extract solids change the specific gravity to the same extent as sucrose. The<br />
EBC method uses Plato's tables while the ASBC method uses Balling's tables<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(Appendix). The extract of the malt is expressed as apercentage (%). This calculation<br />
makes some unwarranted assumptions, so the values are unreliable in absolute terms, but<br />
it gives useful relative values. There are no conversion factors that allow the calculation<br />
of extracts determined by one method to be accurately expressed as the extracts<br />
determinedbyanothermethod,eventhoughtheresultsareroughlycorrelated. Extracts of<br />
pale malts determined by the EBC method are usually in the range 77 83%, (on dry),<br />
whilevaluesfortheIoBmethodareusually300 310lë/kg(ondry).Similarlythetypical<br />
ranges for darker, malts are 75 78% <strong>and</strong> 255 285 lë/kg respectively. Sorghum malts,<br />
which are used to make pale lager-style beers in tropical Africa as well as opaque<br />
African-style beers, are not reliably analysed by the extract determination methods<br />
developed for barley malts, because the gelatinization temperature of sorghum starch is<br />
higher than that of barley (see Table 2.3 on page 38).<br />
Various special, but apparently unst<strong>and</strong>ardized, analytical mashing programmes are in<br />
use (Briggs, 1998). The laboratory mashes differ from brewery mashes in anumber of<br />
important ways. Unlike brewery mashing liquor the water used is distilled <strong>and</strong> contains<br />
no salts, nor is the mash pH adjusted. Also the grist is prepared by using mills that work<br />
differently from brewery mills. The laboratory mashes are dilute compared to brewery<br />
mashes <strong>and</strong>attheendofmashingthegristisnot spargedwithhotwater.Severalattempts<br />
have been made to devise more `brewery-like' laboratory mashes, but they have not been<br />
accepted. Each brewer discovers the relationship between amalt's `lab. extract' <strong>and</strong> the<br />
extract recovered from this malt in the brewery. The extract determinations described<br />
apply to pale malts. Different methods are necessary for special, highly coloured malts<br />
that lack enzymes. For example a50:50 mix of acoloured malt with an enzyme-rich pale<br />
malt may be mashed <strong>and</strong> the extract of the coloured malt is calculated, making the<br />
assumption that the pale malt gives half of the extract it yields when mashed alone.<br />
More information is gained from analysing the mash <strong>and</strong> laboratory worts. The rate of<br />
wort filtration from alaboratory mash does not give agood indication of the brewery<br />
wort run off. `Mashing columns' are needed to achieve this, <strong>and</strong> these devices are not<br />
suitable for routine analyses (e.g., Webster, 1981). Wort colours are determined in<br />
different ways, either visually with colour comparators, or at asingle wavelength, or at<br />
three different wavelengths (the tri-stimulus method, Chapter 19) using a spectrophotometer.<br />
All these approaches have limitations since extracts from different malts not<br />
only differ in colour intensity but also in their spectral characteristics. Because worts<br />
darken to various extents during the hop-boil it is sometimes desirable to measure the<br />
colours of boiled laboratory worts. One problem with the IoB methods is that the wort<br />
from the 450 g mash is more concentrated, <strong>and</strong> therefore more deeply coloured, than that<br />
from the 515 ml mash, <strong>and</strong> so the mashing method employed must be stated when results<br />
are given.<br />
The pH of wort is often routinely recorded. The values obtained vary with the type of<br />
malt. Unusually acid wort (low pH) can be caused by a heavy infestation of microbes on<br />
the malt (Stars et al., 1993).<br />
Malts are analysed for their nitrogen (`protein') contents <strong>and</strong> the laboratory worts are<br />
also analysed for dissolved nitrogenous materials. The values are expressed as nitrogen,<br />
N, in the IoB methods <strong>and</strong> as `protein' (crude protein ˆ 6.25 N) in the EBC <strong>and</strong> ASBC<br />
methods. Because of the differences in the mashing conditions the last two mashes<br />
generally give more soluble nitrogen in the worts than the IoB method. Brewers are<br />
concerned that a malt has a total nitrogen (TN; `protein') content in the specified range<br />
since, outside this range, the wort obtained may differ significantly in quality <strong>and</strong> there<br />
may be problems with extract recovery.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Total soluble nitrogen (TSN) <strong>and</strong> free amino nitrogen (FAN) values are determined.<br />
The TSN needs to be sufficiently high so that the `body' <strong>and</strong> mouth-feel of the beer is<br />
adequate, <strong>and</strong> the beer foam (or `head') will be stable. The soluble nitrogen ratio (SNR;<br />
TSN/TN) of the malt (or the soluble protein ratio or Kolbach Index of the ASBC <strong>and</strong><br />
EBC methods (in each case soluble protein/total protein) serve as measures of<br />
modification <strong>and</strong> avalue is often included in malt specifications. FAN values (chiefly<br />
amino acids <strong>and</strong> small peptides) must be sufficiently high to ensure that lack of<br />
nitrogenous yeast nutrients does not limit fermentation. FAN has been determined by<br />
different methods, which gave different results, so it is essential that the method used is<br />
specified.<br />
Previously,nitrogenousyeastnutrientswereassayedas`formol-nitrogen'.Overtheyears<br />
thepreferrednitrogencontentsforBritishpalealemaltshaverisenfromaround1.3 1.45%<br />
toaround1.65%.Perhapsthisisdueinparttonewervarietiesofbarley<strong>and</strong>changedbrewing<br />
<strong>practice</strong>s. For enzyme-rich malts, needed with high-adjunct brews as in some North<br />
American breweries, TN values of 2.2% (13.8% protein) or more may be preferred. In the<br />
past,othermeasuresweremadesuchaspermanentlysolublenitrogen(PSN)<strong>and</strong>coagulable<br />
nitrogen, respectively the nitrogen in the substances remaining in the wort <strong>and</strong> those<br />
precipitated when the wort was boiled. This method has fallen out of use.<br />
In EBC analysis the time in minutes taken after the mash has reached 70ëC (158ëF)<br />
for samples to stop giving a positive iodine test for starch is recorded as the<br />
`saccharification time'. This is really arough measure of the time taken for the starch to<br />
be dextrinized, <strong>and</strong> is largely dependent on the -amylase content of the malt. The odour<br />
of the mash is noted as, less usually, is the flavour. Both should be normal for the type of<br />
malt being analysed. The appearance of the wort is noted, whether it is clear, opalescent<br />
orturbid.Theactivityofthemixtureofthestarch-degradingenzymesinmaltisestimated<br />
as the `diastatic power', or DP. The enzymes are collectively referred to as diastase. In<br />
principle, soluble starch is incubated with a malt extract <strong>and</strong> the degree of starch<br />
breakdown is estimated after aperiod of incubation at acontrolled temperature. The<br />
results are not highly reproducible <strong>and</strong> represent the joint activities of several enzymes<br />
that are present in different proportions in different malts. Results are expressed in<br />
different units including ëL (degrees Lintner) <strong>and</strong> ëW-K (Windisch-Kolbach units). The<br />
values indicate to brewers if the enzyme content of amalt is adequate.<br />
The level of -amylase in malt extracts is determined by one of several methods. The<br />
level of activity of this enzyme must be adequate if the starch from adjuncts is to be<br />
liquefied in a mash. The ratio of fermentable to non-fermentable sugars is largely<br />
regulated by the activities of the diastatic enzymes during mashing. The fermentabilities<br />
of worts should be constant when brewing a particular beer. Analytically the<br />
fermentabilities of the HWE or Eworts may be determined. However, the fermentability<br />
of these worts increases with storage time as malt enzymes that have survived the<br />
mashing process continue to break down dextrins to simpler, fermentable sugars. Thus<br />
laboratory wort should be boiled to inactivate the enzymes (as occurs in the hop-boil).<br />
Thenitisinoculatedwithapureyeast<strong>and</strong>itisincubatedunderanaerobicconditionsuntil<br />
fermentation is complete. The fall in the specific gravity of the wort allows the<br />
calculation of the attenuation limit of the wort <strong>and</strong> its fermentability (Chapter 4).<br />
When malts contain substantial levels of -glucans modification is incomplete <strong>and</strong><br />
the polysaccharide itself may cause problems in the brewhouse. The -glucans may be<br />
assayed using methods based on enzymes degrading them to glucose, which is measured,<br />
or by the fluorescence of the complex between the polysacchride <strong>and</strong> the reagent<br />
Calcofluor. The activity of the enzymes in malt that degrade -glucans, the -<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
glucanases, are measured either by following the decline in viscosity of a solution of -<br />
glucan incubated with an extract of malt containing the enzyme, or by following the<br />
breakdown of an artificial substrate, a colour-labelled -glucan. Under some<br />
circumstances it is desirable that malts should contain adequate levels of this enzyme,<br />
which is readily inactivated when green malt is kilned using any except the lowest<br />
temperatures. Other substances that may be estimated include the dimethyl sulphide<br />
precursor (DMS-P), which is S-methyl methionine (SMM), <strong>and</strong> N-nitrosodimethylamine<br />
(NDMA). Different styles of beer require different amounts of dimethyl sulphide in the<br />
final product. As the precursor can be destroyed during kilning it is important that its<br />
levels are regulated. NDMA, <strong>and</strong> other less volatile nitrosamines, are suspected of being<br />
carcinogens. When discovered, high levels of NDMA were found in malts, particularly<br />
those from directly fired kilns. However the precautions now taken ensure that the<br />
amounts present are usually below the levels of detection. The levels of NDMA are still<br />
monitored.<br />
The growth of acrospires roughly parallels the advance of modification in malting<br />
grains. The evaluation of acrospire growth in grains in a sample of malt can indicate that<br />
a malt has been made with irregularly germinated material or that good <strong>and</strong> less good<br />
malts have been mixed. In North American <strong>practice</strong> the acrospire lengths of corns in a<br />
sample of malt are classified, by length relative to corn length, as 0 0.25, 0.25 0.50,<br />
0.50 0.75, 0.75 1.0 <strong>and</strong> over 1. Grains in which the acrospire has grown out from<br />
beneath the husk, i.e., over 1.0 in length (so-called overgrown corns, huzzars, cockspurs<br />
or bolters) are undesirable in most kinds of malt as they are deficient in extract.<br />
However, overgrown malts may be relatively rich in enzymes. Specifications may call<br />
for 86 95% of acrospires to fall in the 0.75 1.0 category <strong>and</strong> less than 5% to be<br />
overgrown (1+).<br />
The physical modification of malt grains is traditionally assessed by crushing a series<br />
of corns between finger <strong>and</strong> thumb. Well-modified grains crush to a powder while the<br />
presence of `hard ends' indicates that the apices are not modified <strong>and</strong> completely hard<br />
grains are unmodified. A number of other methods have been used. In the traditional<br />
`sinker test' a h<strong>and</strong>ful of malt corns is thrown into water. Barley corns sink, fully<br />
modified malt corns float horizontally <strong>and</strong> partly modified corns float with the apical<br />
ends downwards. This test is unreliable. The resistance of malt corns to grinding has been<br />
measured, as has the resistance of corns to cutting or penetration by blunt needles.<br />
A device that has achieved wide acceptance is the Friabilimeter. In this a 50 g sample<br />
of malt is broken up between a rotating wire sieve <strong>and</strong> a spring-loaded roller. The friable<br />
material <strong>and</strong> the husk fragments escape through the sieve, <strong>and</strong> the material remaining<br />
after a set period is weighed. The friability is the percentage (by weight) of material that<br />
passes through the sieve. Investigation of material remaining on the sieve can be<br />
informative <strong>and</strong> can indicate if the malt corns generally contain unmodified material or if<br />
a substantial proportion of wholly unmodified grains is present. From this an estimate of<br />
the homogeneity of the malt can be made. Other approaches give indications of the<br />
patterns of modification that have occurred. Samples of malt are stuck to a flat support<br />
<strong>and</strong> a proportion of the grains is ground away with a mechanical s<strong>and</strong>er. In one method<br />
the exposed grain interiors are treated to suppress autofluorescence <strong>and</strong> then they are<br />
treated with Calcofluor. Under UV light this fluoresces when associated with the -<br />
glucans in the unmodified regions of the endosperm cell walls but there is no<br />
fluorescence in the modified regions. Thus the percentage area modified in each exposed<br />
area of endosperm can be assessed, preferably with a specialized automatic scanner. In<br />
the other method about 0.25% of the grains is s<strong>and</strong>ed away <strong>and</strong> then they are exposed to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
an alcoholic solution of the dye methylene blue. This penetrates only into the modified<br />
regions of the grain. The stained grains are dried <strong>and</strong> are s<strong>and</strong>ed further. The blue regions<br />
of the endosperm are modified <strong>and</strong> the white regions are not. Again, the relative areas<br />
modified can be estimated. This data allows the degree of modification <strong>and</strong> its<br />
heterogeneity to be estimated.<br />
The values for `nitrogen modification' (SNR; Kolbach Index) do not always parallel<br />
the estimates of physical modification, <strong>and</strong> indeed the relationships between the two can<br />
differ to an important extent when different varieties of barley are malted in parallel.<br />
Brewers do not want undermodified or overmodified malt. With undermodified malts<br />
extract recoveries in the brewery are unduly low, wort separation can be slow, the worts<br />
obtained may be cloudy, the hot break may form poorly, the wort may have a low<br />
fermentability <strong>and</strong> ferment slowly, the beer may be difficult to filter <strong>and</strong> the filters may<br />
become blocked quickly causing high pressures to build up <strong>and</strong> giving short filtration<br />
runs. Finally the beers may quickly become hazy. In extreme cases -glucan gels may<br />
form <strong>and</strong> deposit. On the other h<strong>and</strong> overmodified malts have their disadvantages. Malt<br />
breakage <strong>and</strong> losses (as dust) are high <strong>and</strong> wort separation may be slowed by the large<br />
proportion of fine particles in the grist. Head retention may be poor, <strong>and</strong> yeast growth can<br />
be wastefully excessive. The hot <strong>and</strong> cold breaks may be heavy, the wort may contain<br />
finely divided material that is hard to remove by filtration <strong>and</strong>, because of the excessive<br />
levels of reducing sugars <strong>and</strong> amino acids present, the wort may darken too much on<br />
boiling, due to the formation of melanoidins.<br />
From time to time other analyses may be performed. Thus levels of arsenic, lead,<br />
cadmium <strong>and</strong> iron may be determined to check for the absence of contamination. Zinc<br />
can be measured but, as there is no clear relationship between this <strong>and</strong> the amount of zinc<br />
available to the yeast in the wort, this is rarely done. Levels of microbes, especially<br />
Fusaria, may be determined <strong>and</strong> several tests for agents causing gushing have been<br />
devised (Donhauser et al., 1991; Vaag et al., 1993). Sometimes samples may be analysed<br />
for halogenated contaminants (such as chlorinated substances), or residues of insecticides<br />
or agricultural chemicals. Where the use of added gibberellic acid is forbidden residues of<br />
this substance may be sought on the malt's surface. As this substance occurs naturally<br />
within the grains the results of such tests must be suspect.<br />
2.2.5 Types of kilned malt<br />
Malt types are not clearly distinct. The descriptions given here are representative.<br />
Different breweries specify distinctly different malts giving them the same title (`pale<br />
lager', mild, ale, etc.). The question of what constitutes a sensible malt specification is<br />
discussed later. Extensive sets of malt analyses are available (Briggs, 1998; Narziss,<br />
1976; 1991). In this chapter malts are described with emphasis on the aspects most of<br />
interest to brewers. Where the use of adjuncts is forbidden, as by the German<br />
Reinheitsgebot, chit malts <strong>and</strong> short grown malts may be used. These are less expensive<br />
to produce than `normal' malts. They retain some raw grain characteristics <strong>and</strong> have<br />
some of the advantages that are gained from using unmalted grains as adjuncts. These<br />
malts are made by steeping barley to a low moisture content <strong>and</strong> then, either as soon as<br />
the grain has chitted or after a short period of germination, the `green malt' is dried at a<br />
low temperature. The malting losses occurring in making these materials are small <strong>and</strong>,<br />
because of their low moisture contents, they are comparatively inexpensive to kiln. The<br />
products have moisture contents of 2 5% <strong>and</strong> contain some hydrolytic enzymes but the<br />
endosperms are incompletely modified. Their Kolbach indices are low <strong>and</strong> extracts may<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
e 77 80% (on dry, EBC) with fine-coarse extract differences of 6 12%. These malts<br />
provide less expensive extract <strong>and</strong> better beer foam stability than conventional malts, but<br />
they enhance wort <strong>and</strong> beer viscosity, slow wort separation <strong>and</strong> reduce the rate of beer<br />
filtration <strong>and</strong> the length of filter runs. Short grown green malts have been flaked before<br />
use, without being kilned, which facilitates extract recovery but destroys the enzymes<br />
originally present.<br />
Kilning is expensive, so attempts have been made to brew with green, unkilned malt.<br />
This material is unstable <strong>and</strong> must be used as soon as it is ready. It is exceptionally rich in<br />
enzymes <strong>and</strong> yields highly fermentable, proanthocyanidin-poor wort with a good extract.<br />
Its high content of hydrolytic enzymes allow it to convert a high proportion of adjuncts in<br />
mashes (Briggs et al., 1981). This material has rarely been used both because of its<br />
instability <strong>and</strong> because it imparts unpleasant flavours in the finished beers. Roots remain<br />
attached to green malts. A compromise would be to have malts dried at a low temperature<br />
to 7 8% moisture. Roots can be removed from such material, which can be stored for<br />
some weeks. It contains high levels of hydrolytic enzymes <strong>and</strong>, because less intensive<br />
drying is needed on the kiln, is less expensive to produce than normally kilned malt. Such<br />
material seems not to give unwanted flavours to beers. Despite these advantages such<br />
malt is apparently not in use.<br />
Lager beers are widely produced, <strong>and</strong> all malts making these beers are, by definition,<br />
lager malts. In Germany, where most beers are lagers, the palest to the darkest malts are<br />
lager malts. In the UK lager malts <strong>and</strong> lager beers are all pale. In North America the<br />
lager-style beers are usually brewed using high levels of unmalted adjuncts in the mash<br />
<strong>and</strong> so the malts used differ significantly from European lager malts. The characteristics<br />
of European lager malts have changed during the last century, <strong>and</strong> the differences<br />
between British pale ale malts <strong>and</strong> lager malts have become indistinct. North American<br />
malts may be made from two- or six-row barleys or mixtures of both. In general their<br />
nitrogen contents are high (TN 1.7 2.3%; 10.6 14.4% protein), with soluble to total<br />
protein ratios of 43 48% <strong>and</strong> high levels of FAN, <strong>and</strong> hydrolytic enzymes (DU values of<br />
30 45 or even 50 for -amylase), all characteristics needed with adjunct-rich mashes.<br />
The high nitrogen contents are associated with lower extracts (77 81% on dry) but this is<br />
of less significance when so much of the extract is derived from adjuncts. Such malts are<br />
pale (1.4 1.9 ëLovibond) <strong>and</strong> have moisture contents of 3.7 4.3%.<br />
The palest of the European products are Pilsen malts (Pilsener Malz). In the past these<br />
were undermodified but now they are fully modified <strong>and</strong> are prepared from barleys<br />
having moderate nitrogen contents. They are kilned at low temperatures to minimize<br />
colour formation. Typical analyses are E, at least 81% (EBC, on dry), fine-coarse extract<br />
difference 1 2%; TN, 1.68 (10.5% protein); Kolbach index 38 42%; moisture less than<br />
4.5%; -amylase 40 DU; DP 240 300 ëW.-K.; saccharification time 10 15 min.; colour,<br />
2.5 3.4 ëEBC; boiled wort colour, 4.2 6.2 ëEBC; wort pH, 5.9 6.0. Helles (pale; light)<br />
malts are rather similar, but are made from barleys richer in nitrogen. British lager malts<br />
are all pale <strong>and</strong> well modified. Analyses are usually in the ranges: HWE 300 310 lë/kg<br />
(on dry), TN, 1.55 1.75%; TSN, 0.5 0.7%; SNR, 31 41%; DP, not more than 70 ëIoB;<br />
moisture less than 4.5%; saccharification time less that 15 minutes. Colour may be<br />
3.0 ëEBC. Because of the low temperatures used in kilning lager malts (finishing curing at<br />
e.g., 70 ëC; 158 ëF) are rich in enzymes <strong>and</strong> so sometimes give slightly higher extracts<br />
than pale ale malts, which are cured at higher temperatures (finishing at 95 105 ëC;<br />
203 221 ëF), <strong>and</strong> have more characteristic flavours but lower enzyme activities.<br />
In the last 50 60 years the moisture contents of the best pale ale malts have been<br />
allowed to rise from 1.5% to a maximum of 3% <strong>and</strong> preferred TN values have increased<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
from as little as 1.35% to around 1.65%. Other preferred current analyses are: HWE,<br />
306 310 lë/kg (on dry); TN from less than 1.55% to less that 1.70%; TSN, 0.5 0.7%;<br />
SNR, 31 42%; FAN, 0.1 0.12%; CWE, 18 22%, colour 4 6 ëEBC; <strong>and</strong> DP,<br />
44 65 ëIoB. The malt will have a high friability. Many ale <strong>and</strong> lager malts are made<br />
to meet customers' particular specifications. However, it is more economical for<br />
maltsters to make larger volumes of `st<strong>and</strong>ard' malt having specifications close to those<br />
of many of their customers <strong>and</strong> offer these at lower prices. The `st<strong>and</strong>ard malts' from<br />
different suppliers have different specifications <strong>and</strong> these are likely to change when<br />
barley quality changes. Mild ale malts are generally made from lower-quality barleys <strong>and</strong><br />
will be slightly less well modified <strong>and</strong> will be more strongly coloured than pale ale malts.<br />
A representative mild malt might have a HWE of 305 lë/kg (on dry), <strong>and</strong> a fine-coarse<br />
extract difference of about 5 lë/kg; a moisture content of 3.5%; CWE, 18 21%; TN,<br />
1.55 1.75%; TSN, 0.6 0.7%; SNR, 36 40%. DP, 40 60; colour, 6 9 ëEBC. Thus in<br />
UK malts the colours increase as one progresses from lager to pale ale to mild.<br />
In German <strong>practice</strong> the next type is Viennese malt (Wiener Malz), which is used for<br />
making `golden' lagers. This is made from normally modified green malt kilned to a final<br />
temperature of about 90 ëC (194 ëF), giving a colour of 5.5 6.0 ëEBC. Munich malt<br />
(MuÈnchener Malz) is relatively dark, very well modified <strong>and</strong> aromatic <strong>and</strong> is made by<br />
germinating nitrogen-rich barley, steeped to a high moisture content, so that it is well<br />
grown (all acrospires at least three-quarters grown) <strong>and</strong> finishing germination warm, at<br />
25 ëC (77 ëF). Kilning involves some stewing <strong>and</strong> curing is finished at 100 105 ëC<br />
(212 221 ëF), conditions causing appreciable enzyme destruction. This malt has a colour<br />
of 15 25 ëEBC. The wort is rich in melanoidin precursors <strong>and</strong> darkens on boiling, e.g.,<br />
from 15 to 25 ëEBC. Other typical analyses are: E, 80%, (on dry); fine-coarse extract<br />
difference 2 3%; total protein 11.5% (TN, 1.84%); Kolbach index, 38 40%;<br />
saccharification time, 20 30 min.; wort fermentability, about 75% (compared to wort<br />
from Pilsen malt of about 81%). -Amylase <strong>and</strong> DP values are low, at 30 DU <strong>and</strong><br />
140 ëW.-K. respectively. Analyses of a British made, Munich-style malt are: HWE, 300<br />
lë/kg, (on dry); moisture 4.5%; TN, less than 1.65%, TSN less than 0.65%, colour about<br />
15 ëEBC <strong>and</strong> DP at least 30 ëIoB.<br />
Brumalt (BruÈhmalz) is an even darker German malt, which is made by steeping a<br />
nitrogen-rich barley (e.g., TN 1.84%; protein 11.5%) <strong>and</strong> germinating it at an<br />
exceptionally high moisture content, about 48%. When the grain is well grown it is<br />
held in a closed container for, say, 36 hours so that the oxygen is used up <strong>and</strong> carbon<br />
dioxide accumulates. The temperature rises to 40 50 ëC (104 122 ëF) <strong>and</strong> the grain<br />
contents soften <strong>and</strong> become pulpy <strong>and</strong> rich in reducing sugars <strong>and</strong> amino acids. These<br />
melanoidin precursors interact when the green malt is kilned, at 80 90 ëC (176 194 ëF),<br />
to give a highly aromatic <strong>and</strong> melanoidin-rich material with a high colour, usually of<br />
30 40 ëEBC. Such malt gives characteristic rich flavours to beers <strong>and</strong> these are said to be<br />
stabilized by the reductones from the malt. Sometimes this general kind of material is<br />
called rH malt or melanoidin malt.<br />
Some malts are made from barleys lacking proanthocyanidins (anthocyanogens;<br />
Briggs, 1998; Sole, 2000). The absence of the polyphenolic haze precursors means that<br />
beers made from these malts <strong>and</strong> using polyphenol-free hop extracts are unlikely to<br />
become hazy. Other `conventional' malts are in use. Some special beers are brewed using<br />
a proportion of smoked malt (Rauch Malz) to gain a `smoky' flavour, but these contain<br />
elevated levels of NDMA. Such flavours were once common, when malts were all kilned<br />
using direct-fired, wood-burning kilns. Now the smoke from a wood-burning furnace is<br />
led into the hot stream of air entering the bed of green malt on the kiln.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Acid- or lactic-malts have been used sporadically in the UK <strong>and</strong> more frequently in<br />
Germany to adjust the mash pH. Originally they were used to offset the effects of<br />
bicarbonate-rich mashing liquor. These malts carry 2 4% of non-volatile lactic acid.<br />
They are made in several different ways, for example by steeping or spraying green malt<br />
with solutions of biologically prepared lactic acid during germination <strong>and</strong> before kilning.<br />
Application of lactic acid to germinating green malt can check rootlet growth <strong>and</strong> the rise<br />
in malting losses <strong>and</strong> favour the accumulation of soluble nitrogenous substances. In<br />
another system pale malt is steeped in water at 45 47 ëC (113 116.6 ëF) for an extended<br />
period. Sugars are leached into solution <strong>and</strong> thermophilic lactic acid bacteria convert<br />
much of them to lactic acid. The grain is re-dried, <strong>and</strong> the acidic steeping water is reused.<br />
The use of such malts (as 5 10% of the grist) lowers mash pH <strong>and</strong>, at least with<br />
some of these malts, the CWE is increased as is the HWE, the TSN <strong>and</strong> the FAN. A<br />
typical analysis (IoB) is HWE, over 297 lë/kg (on dry); moisture, 5 6%; TSN,<br />
0.8 1.2%; colour, 10 25 ëEBC; lactic acid, 2.1 2.5%. Such malts are often used where<br />
the addition of chemically prepared acids for pH adjustment is forbidden.<br />
Some brewers prefer to add lactic acid, prepared biologically from wort (using<br />
Lactobacillus delbruÈckii), for mash pH adjustment. Various other materials have been<br />
added to malts. For example, added formaldehyde was not readily detectable, but beers<br />
made from such malt were low in proanthocyanidins <strong>and</strong> unlikely to form non-biological<br />
haze. In the Belmalt process green malt was sprayed with a solution of glucose syrup<br />
(3.5 kg/100 kg original barley) some hours before kilning. The malts were more acid <strong>and</strong><br />
gave higher extracts <strong>and</strong> levels of soluble nitrogen than controls <strong>and</strong> the worts had higher<br />
attenuation limits. Residual glucose would account for the higher extract <strong>and</strong><br />
fermentability <strong>and</strong> the conversion of some of the glucose to lactic acid by bacteria on<br />
the grain would cause mash acidification <strong>and</strong> so increase the TSN. Gum arabic has been<br />
sprayed onto malts to increase the head retention of beers made from them. Solutions of<br />
thermostable enzymes, probably including amylase <strong>and</strong> -glucanase, have been applied<br />
to malts, presumably to enhance their apparent quality. These enzymes will not penetrate<br />
into the interiors of grains <strong>and</strong> so must exert their effects when carried forward into the<br />
mash. Generally it is better for brewers to keep these kinds of additions under their own<br />
control <strong>and</strong> make them at mashing or later in brewing.<br />
Malts are made from cereals other than barley (Briggs, 1998; Byrne et al., 1993;<br />
Narziss, 1976; Taylor <strong>and</strong> Boxall, 1999). Wheat malts are generally pale, although dark<br />
wheat malts are `made to order'. In mainl<strong>and</strong> Europe wheat malts make up the major<br />
parts of the grists (up to 80%) of special beers, including the German top-fermented<br />
Weissbier (white beer) <strong>and</strong> Weizenbier (wheat beer). In the UK small amounts of wheat<br />
malt (3 10%) may be included in grists to improve the foam formation <strong>and</strong> head<br />
retention of the beers. Other benefits claimed are improved beer clarity <strong>and</strong> palatefullness.<br />
The flavour of wheat extract is relatively `neutral'. Wheat malts tend to be<br />
undermodified <strong>and</strong> their inclusion in the mash can lead to slower wort run off <strong>and</strong><br />
sometimes fining problems in beers. Wheat has a naked grain, so it is easily damaged<br />
during h<strong>and</strong>ling <strong>and</strong> the acrospire (coleoptile) is not protected by a husk during<br />
germination. Water uptake is rapid during steeping. Usually soft wheats (TN less than<br />
1.9%) are malted <strong>and</strong>, like barley, they will respond to applications of gibberellic acid.<br />
Because of the absence of husk, which yields no extract, extracts of wheat malts can be<br />
relatively high, e.g., 328 lë/kg (on dry), 86%. Wheat malts may have moisture contents of<br />
5%, colours of 2 6 ëEBC; a TN of 1.87%, an SNR of 38 40%, a Kolbach index of up to<br />
or over 50%, <strong>and</strong> high levels of diastatic enzymes. Rye varieties have thin, naked grains<br />
<strong>and</strong> the malts made from them can confer unusual flavours (toffee, caramel) to beers,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
with changes to the mouth-feel (palate; smoother; more mellow) <strong>and</strong> a slight<br />
improvement in head retention. Rye malt may give ared tinge to beer. Rye varieties<br />
differ greatly in their suitability for making malts. Compared to barley the thin, naked<br />
grain takes up water quickly. Pale rye malts (2 8ëEBC) have exceptionally high but<br />
variable extracts (about 315 lë/kg, on dry, even 85 90% EBC) <strong>and</strong> very varied nitrogen<br />
contents (1.2 2.3%) <strong>and</strong> varied levels of diastatic enzymes. The malts are only<br />
occasionally used in special beers.<br />
Malts have been made from the `synthetic' wheat-rye hybrid triticale (Triticosecale)<br />
which, like its parents, has naked grains <strong>and</strong> takes up water rapidly during steeping.<br />
There are wide varietal differences in the malting quality of triticale cultivars. Although<br />
extracts can be very high (values of 335 lë/kg, (on dry) <strong>and</strong> 88 90% EBC have been<br />
reported) yet triticale malts are not used, possibly because the grain has atendency to<br />
have high TN values <strong>and</strong> the malts yield hazy worts rich in soluble nitrogen, including<br />
suspended <strong>and</strong> finely divided protein largely derived from prolamines. The worts are<br />
very viscous because of dissolved pentosans (Blanchflower <strong>and</strong> Briggs, 1991; Byrne et<br />
al., 1993). When milled, triticale malts give rise to afine flour which impedes wort<br />
separation. In the past, stouts were made with high proportions of oat malts in the grists.<br />
The reasons for using oat malts are not clear <strong>and</strong> fermentation problems (foaming<br />
fermentations <strong>and</strong> cloudy worts) were often encountered. However, the high husk<br />
content of oats meant that the `husky' grist favoured wort run off at the end of mashing.<br />
Small amounts of oat malts are now used to give character to special beers (toasted,<br />
biscuit-like aroma <strong>and</strong> an intense mouth-feel). The malts have some unusual<br />
characteristics. Extracts are low (e.g., 230 234 lë/kg on dry), the lipid content of the<br />
malts is high with the danger of the material becoming rancid <strong>and</strong> the beer having poor<br />
head retention <strong>and</strong> flavour instability. The TN value may be moderate (e.g., 1.6%), <strong>and</strong><br />
the SNR is low (e.g., 18%). Diastatic power may be about the same as that of abarley<br />
malt having asimilar nitrogen content.<br />
Although malts are made from several tropical cereals only those made from sorghums<br />
have attracted much attention. In places in Africa millets have been malted mixed with<br />
sorghum grain, probably for the extra starch degrading enzymes provided by the malted<br />
millet (Chapter 16). Sorghum grain is steeped <strong>and</strong> watered during germination. Often the<br />
grain is treated with substances such as formaldehyde or sodium hydroxide in attempts to<br />
control the surface microbes. Relative to barley, sorghum is malted warm (e.g., 25 ëC; 77 ëF).<br />
The rootlets <strong>and</strong> the shoots are removed from the dried malt when lager-style beer is being<br />
made but not when the malt is for making opaque beer. The seedling tissues are rich in<br />
hydrogen cyanide. The information available on malted sorghum is inconsistent. This is<br />
probably because malted sorghum is used in making opaque beers (Chapter 16) as well as<br />
lagers <strong>and</strong> the requirements for these processes are different because of the large differences<br />
between different varieties of sorghum <strong>and</strong> because the methods of analysis <strong>and</strong> mashing<br />
that are applicable to barley malts are not all suitable for sorghum malts. This is primarily<br />
because the gelatinization temperatures of sorghum starches are higher than the starches of<br />
barley or wheat malts. Thus the extract of a sorghum malt determined using an inappropriate<br />
method (mashing at 65 ëC; 149 ëF) was found to be 112 lë/kg while mashing in an appropriate<br />
way gave an extract of 268 lë/kg. Elsewhere extracts from sorghum malts of between 65 <strong>and</strong><br />
85% were found, with fine-coarse extract differences in the range 0.5 18.2%. Estimates of<br />
other analyses also varied widely. It has been generally accepted that sorghum malts are<br />
deficient in enzymes (this is not a problem when making traditional, opaque beers), but this<br />
may not be not true for malts made from some varieties.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
2.2.6 Special malts<br />
The malts already described are all `finished' on kilns. However, there is a group of malts<br />
which are finished in roasting drums (Bemment, 1985; Briggs, 1998; Gretenhart, 1997;<br />
Maule, 1998; Narziss, 1976). All these special malts are used as small proportions of<br />
grists to give particular colours, flavours <strong>and</strong> aromas (i.e., to impart characters) to beers.<br />
They can be considered in two groups; those that are prepared by a simple heating<br />
process, such as amber, diamber, brown, chocolate <strong>and</strong> black malts (<strong>and</strong>, by tradition in<br />
the UK, roasted barley), <strong>and</strong> crystal <strong>and</strong> caramel malts in which the wet malts are<br />
`stewed' so that the endosperm contents are liquefied before they are dried <strong>and</strong> cooked. In<br />
each group a wide range of colours occurs. As the colour ranges are continuous <strong>and</strong> as the<br />
qualities of the starting materials can be varied as, to some extent, can the roasting<br />
regimes, it follows that the number of malts that might be made is unlimited. The more<br />
usual types <strong>and</strong> divisions are described here, but intermediate types can be made <strong>and</strong><br />
sometimes are. Because these malts are required primarily for the characters <strong>and</strong> colours<br />
that they provide, extract <strong>and</strong> colour are the analyses which, together with moisture<br />
content, are usually specified. Sometimes coloured malts are made from wheat or other<br />
cereals but only the barley malts are in common use. Unlike `white' malts, coloured malts<br />
should be used as fresh as possible, storage time being minimized, to retain their aromas.<br />
During their preparation the heating is so intense that no enzymes survive. As the colour<br />
increases in a series of malts so the malt extracts decline slightly as the extra colour is<br />
generated by more extreme or more prolonged heating. For example, in a series of<br />
German caramel malts the extracts <strong>and</strong> colours were: Carapils, E 78%, colour,<br />
2 5 ëEBC; Carahell, E 77%, colour 20 25ë EBC; CaramuÈnch, E 76%, colour<br />
50 300 ëEBC. As the colour increases so the wort pH values tend to decrease <strong>and</strong> the<br />
Kolbach indices decline.<br />
Amber malts are prepared by roasting pale ale or mild malts or, after drying, well<br />
modified green malts. Heating programmes begin at about 48 ëC (118.4 ëF) <strong>and</strong> rise to<br />
about 170 ëC (338 ëF). The `normal' colour range varies, but is usually 40 85 ëEBC.<br />
Moisture contents are 3.5% or less. Extracts vary between 270 <strong>and</strong> 285 lë/kg. These malts<br />
are valued for giving characteristic dry palates <strong>and</strong> baked or biscuit-like flavours to<br />
golden-coloured ales. Diamber malts are probably not made now, but modern brown<br />
malts are similar to amber malts prepared at higher temperatures. Such malts may have<br />
extracts of 260 280 lë/kg <strong>and</strong> colours of 90 150 ëEBC. Chocolate <strong>and</strong> black malts <strong>and</strong><br />
roasted barley are also prepared in roasting cylinders but, relative to amber <strong>and</strong> brown<br />
malts, the heating is much more severe <strong>and</strong> there is a risk that the grain may catch fire.<br />
The process must be regulated so that no charring occurs <strong>and</strong> that when cut the grains are<br />
evenly coloured <strong>and</strong> have a floury texture, with no glassiness, <strong>and</strong> have the correct colour<br />
throughout. Well-modified green malt (TN 1.5 1.7%) is carefully dried <strong>and</strong> dressed. The<br />
material is loaded into a roasting cylinder <strong>and</strong> the temperature is increased from about<br />
75 ëC (167 ëF) to 175 ëC (347 ëF) <strong>and</strong> then more slowly to 215 ëC (419 ëF) for chocolate<br />
malts <strong>and</strong> to 225 ëC (437 ëF) for black malts. During roasting, unpleasant fumes are<br />
released <strong>and</strong> these must be eliminated by scrubbers or after-burners. Roasted barley is<br />
finished at a higher temperature, 230 ëC (446 ëF). Towards the end of roasting, when the<br />
heaters are switched off, the temperature of the load continues to rise as heat is generated<br />
in the grain. At this stage the operator checks colour every 2 3 min. <strong>and</strong> at the correct<br />
moment quenches the load with a spray of water.<br />
Roasted barleys usually have colours in the range 1200 1500 ëEBC. These are used in<br />
making some stouts <strong>and</strong> impart `sharp', `dry', `acidic' or `burnt' notes to the product. In<br />
contrast to roasted malts roasted barley gives no hint of sweetness. The roasted grains<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
should be reddish-black, shiny <strong>and</strong> swollen <strong>and</strong> a proportion will be split. Extracts are<br />
HWE, 260 275 lë/kg, <strong>and</strong> moisture contents will be less than 2%. Pale chocolate malts<br />
will have colours of 500 600 ëEBC, <strong>and</strong> the more usual chocolate malts<br />
900 1100 ëEBC. Black malts have colours of 1200 1400 ëEBC. All have moisture<br />
contents of 2% or less <strong>and</strong> extracts of 255 275 lë/kg. Flavour descriptions of these<br />
materials are not satisfactory, but they include `dry', `burnt', `acid' <strong>and</strong> `astringent' but<br />
when chewed they have a residual sweetness which is distinct from the flavour of roast<br />
barley. The hot water extracts of chocolate <strong>and</strong> roasted malts <strong>and</strong> roasted barley are<br />
determined on finely ground samples mashed with boiling water at 100 ëC (212 ëF) in the<br />
IoB method, so enzymolysis is not involved.<br />
Crystal <strong>and</strong> caramel malts are unique in that during their preparation the endosperm<br />
contents are deliberately mashed, stewed <strong>and</strong> liquefied <strong>and</strong>, when cut, the finished malts<br />
should be hard <strong>and</strong> all the grains should be glassy or `crystalline' in appearance. These<br />
malts are prepared in a wide range of colours, some of which are named. They impart rich<br />
<strong>and</strong> delicious <strong>and</strong> other characteristic flavours <strong>and</strong> they give body to beers <strong>and</strong> are<br />
believed to improve beer stability. Sound barley, sometimes with a high nitrogen content<br />
of 1.7 2.0%, is malted. When it is well modified either the green malt is taken to a<br />
roasting drum directly or, less economically, it is lightly kiln dried. The green malt, or the<br />
re-wetted, kilned malt is warmed <strong>and</strong> held moist at a temperature of 60 75 ëC<br />
(140 167 ëF) until the contents of the grains are liquefied <strong>and</strong> the liquid contents can be<br />
squeezed out. The temperature is then increased <strong>and</strong> the grain is ventilated with hot air so<br />
that both cooking <strong>and</strong> drying occur. The liquefaction step ensures that starch, <strong>and</strong><br />
possibly the endosperm cell-walls, are degraded <strong>and</strong> sugars <strong>and</strong> other soluble materials<br />
accumulate. Thus on heating <strong>and</strong> drying <strong>and</strong> depending on the exact conditions a<br />
concentrated sugar solution is produced together with various amounts of melanoidins<br />
<strong>and</strong> flavour <strong>and</strong> aroma substances.<br />
The finished product is rapidly cooled, <strong>and</strong> the contents solidify to a sugary, solid<br />
mass. Moisture contents are 3 7.5% <strong>and</strong> extracts are 260 285 lë/kg; 76 80%. Preferred<br />
colour ranges are about 20, 120 140 <strong>and</strong> 300 500 ëEBC. While the palest crystal malts<br />
are sweet, the darker malts have more complex flavours with caramel-, toffee-, malty-,<br />
aromatic-, honey-like <strong>and</strong> luscious characters becoming more apparent until in the<br />
darkest products harsher, burnt flavours appear. These products are variously called<br />
caramel or crystal malts. It has been shown that the flavour spectra can usefully be varied<br />
(Ch<strong>and</strong>ra et al., 1999).<br />
2.2.7 Malt specifications<br />
When brewers purchase malt they require it to be excellent in quality <strong>and</strong> moderate in<br />
price. They expect the extract yield <strong>and</strong> quality will be good <strong>and</strong> that beer production will<br />
run smoothly <strong>and</strong> yield a good product. Malts have different properties <strong>and</strong> are used to<br />
produce different types of beer. Brewers need to decide what analyses define the best<br />
malt with which to make a particular beer, <strong>and</strong> to agree with maltsters that this is what<br />
can <strong>and</strong> will be supplied. The analyses available do not reliably predict a malt's<br />
brewhouse performance <strong>and</strong> brewers have yet to agree on what set of analyses should be<br />
used to specifically define a malt. Cheap, poorly made malts are often undermodified<br />
<strong>and</strong>/or inhomogeneous <strong>and</strong> brewing with them can give rise to costs resembling those<br />
arising from mashing with excessive levels of particular adjuncts. For example, failure to<br />
recover the expected extract in the brewhouse or the need for a lengthened mashing<br />
programme, slow wort separation prolonging the lautering stage <strong>and</strong> so disrupting the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
production programme, excessive break-formation in the hop-boil, short filter runs <strong>and</strong><br />
slow beer-filtration rates so the production cycle is further disrupted <strong>and</strong> extra filter aids,<br />
e.g., kieselguhr (diatomaceous earth) may be needed. Furthermore, there may be a need<br />
to use extra beer stabilization treatments <strong>and</strong>/or add extra enzymes to the mash.<br />
Inadequate yields of small nitrogenous molecules (FAN), that then limit yeast growth <strong>and</strong><br />
fermentation, may also occur as may low carbohydrate fermentability (too few<br />
fermentable sugars) that ensures that alcohol yield is depressed. All these problems<br />
cause disruption in the production schedule <strong>and</strong> increase costs.<br />
In <strong>practice</strong>, brewers use different analyses in attempts to ensure that malts meet their<br />
requirements <strong>and</strong> the situation is complicated by the ongoing search for <strong>and</strong> introduction<br />
of improved methods (Aastrup et al., 1991; Briggs, 1998; Buckee, 1997; Copestake,<br />
1998; Gromus, 1988; Hyde <strong>and</strong> Brookes, 1978; Seward, 1992). The brewer may specify<br />
the variety(ies) of barley from which the malt may be made, <strong>and</strong> the harvest year,<br />
whether or not abrasion <strong>and</strong>/or additives may be used, details of the kilning cycle, <strong>and</strong> a<br />
minimum (or maximum for coloured malts) period between manufacture <strong>and</strong> delivery. A<br />
specification will contain an upper limit to screenings (thin corns) <strong>and</strong> dust, a maximum<br />
moisture content, a measure of the laboratory extract coupled to a lower limit, sometimes<br />
a preferred range for the fine-coarse extract difference, a total nitrogen (protein) limit or<br />
range, a range or limit for the total soluble nitrogen (protein) value <strong>and</strong> for the SNR or<br />
Kolbach Index, <strong>and</strong> often a lower limit for the free amino nitrogen. Values (maximum,<br />
minimum or ranges) may be specified for DP, -amylase <strong>and</strong> saccharification time, <strong>and</strong><br />
limits may be set on the concentration of the DMS precursor.<br />
In addition, an upper limit or a range will be set for the colour of the laboratory wort,<br />
often before <strong>and</strong>/or after boiling. To these may be added specified limits for the other<br />
characteristics of the laboratory wort, including smell, clarity, pH, viscosity <strong>and</strong> -glucan<br />
content, <strong>and</strong> estimates of malt -glucanase, friability, homogeneity, <strong>and</strong> any others,<br />
including wort fermentability. As many of these values can be determined in more that<br />
one way <strong>and</strong> the results of analyses may be expressed in non-st<strong>and</strong>ard ways, even<br />
including non-st<strong>and</strong>ard units, a maltster in international trade may need to recognize<br />
nearly 300 analytical values, a situation so bizarre as to be ridiculous. In addition a<br />
guarantee may be needed to indicate that the malt is not contaminated with lead, arsenic<br />
or nitrosamines, mycotoxins or unapproved agricultural chemicals, insecticides or<br />
fumigants.<br />
Two other kinds of problem arise. The first relates to the brewer specifying<br />
combinations of malt characteristics that cannot be combined in one product. For<br />
example, it is not possible to produce a pale malt with a very rich flavour, or an enzyme<br />
rich malt that has a high colour. Malts with low SNRs cannot be made highly friable.<br />
Barleys with low nitrogen (protein) contents cannot be malted to give products<br />
exceptionally rich in enzymes, high nitrogen contents cannot be combined with high<br />
carbohydrate extracts. Malts with poor physical modification cannot have low -glucan<br />
contents, <strong>and</strong> so on. These facts are the inevitable consequences of the composition of<br />
barley <strong>and</strong> the integrated way in which changes in the grain occur during malting. The<br />
second kind of difficulty arises from drawing up specifications that are too inflexible, or<br />
`tight', so that they cannot be met routinely. For example, it is ridiculous to specify a<br />
particular analytical value without taking account of analytical variations <strong>and</strong> the<br />
variations that occur in barley. It is meaningless to specify that a malt's nitrogen content,<br />
TN, must be 1.65% exactly, when the repeatability <strong>and</strong> reproducibility values for the<br />
analysis are 0.049% <strong>and</strong> 0.085% respectively according to the Recommended Methods of<br />
the Institute of <strong>Brewing</strong>. Realistic specifications must be agreed between maltsters <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ewers, probably annually, taking into account the changing varieties of barley being<br />
grown <strong>and</strong> the quality of the barleys available from the current harvest.<br />
2.3 Adjuncts<br />
Adjuncts are materials, other than malt, that are sources of extract (Briggs, 1998; Byrne<br />
<strong>and</strong> Letters, 1992; Letters, 1990; Lloyd, 1986, 1988a, b; Martin, 1978; Stowell, 1985).<br />
They are used because they yield less expensive extract than malt <strong>and</strong>/or they impart<br />
desirablecharacteristicstotheproduct.Forexample,theymaydilutethelevelsofsoluble<br />
nitrogen <strong>and</strong> polyphenolic tannins in the wort, allowing the use of high-nitrogen (protein<br />
rich) malts <strong>and</strong> the production of beer less prone to form haze. Some adjuncts enhance<br />
head formation <strong>and</strong> retention. The higher proportion of adjuncts used in amash the more<br />
difficult it is to achieve good extract recoveries <strong>and</strong> also wort viscosity is often increased,<br />
run off is slowed <strong>and</strong> fermentability is reduced. The addition of soluble sugars or syrups<br />
to the wort effectively increases the capacity of the brewhouse <strong>and</strong> provides asimple<br />
method for generating high-gravity worts <strong>and</strong> adjusting wort fermentability. Solid, `mash<br />
tun' adjuncts may be added to the grist <strong>and</strong> the starch they contain will be hydrolysed by<br />
enzymes from the malt or from other sources (Section 2.5). Other soluble preparations,<br />
sugars <strong>and</strong> syrups, otherwise `copper'- or `kettle'-adjuncts, are dissolved in the wort<br />
during the hop-boil. In addition to these abrewer may add other sugars to the beer as<br />
`primings', <strong>and</strong> caramels or other materials may be added to adjust beer colour.<br />
Adjuncts are analysed according to the official sets of methods (Section 1.15.1, p. 9).<br />
Since, like special malts, mash tun adjuncts are largely or wholly lacking in hydrolytic<br />
enzymes, analytical mashes are made with adjunct, which may be pre-cooked <strong>and</strong> mixed<br />
50:50 with an enzyme-rich malt. The extract yield of the adjunct is calculated, assuming that<br />
the extract recovered from the malt is half that which is obtained from an all-malt mash. The<br />
copper adjuncts are dissolved <strong>and</strong> the characteristics of the solutions are determined. The<br />
analytical values of interest include the specific gravity <strong>and</strong> hence the extract, the colour <strong>and</strong><br />
clarity of the wort (often the colour is very low), the yield of soluble nitrogen, the oil content,<br />
the sulphur dioxide content, the pH, the ash content, levels of heavy metals (such as iron,<br />
copper, lead <strong>and</strong> arsenic), the level of microbes, the spectrum of sugars present <strong>and</strong> the<br />
fermentability of the mixture, the flavour, aroma <strong>and</strong> purity of the material <strong>and</strong> the absence<br />
of any deterioration. The amounts of adjuncts used vary widely. In some places their use is<br />
forbidden. In North America 60% of the extract in a brew may be derived from adjuncts,<br />
while elsewhere 10 20% is more usual. It is feasible to make beers with up to 95% of the<br />
grist being raw barley (Briggs et al., 1981; Wieg, 1973, 1987).<br />
The choice of adjunct(s) requires care. The material chosen must be regularly<br />
available in adequate amounts <strong>and</strong> be of good quality. The use of this material must<br />
enhance, or at least not reduce, the quality of the beer being made. It is difficult to switch<br />
between different kinds of adjunct. Apart from the risk of altering the nature of the beer,<br />
changing adjuncts may require alterations in the brewery equipment. For example, the<br />
h<strong>and</strong>ling plant needed for syrups is completely different from that needed for any mash<br />
tun adjunct <strong>and</strong> the equipments needed to h<strong>and</strong>le flours, flakes <strong>and</strong> grits are all different.<br />
2.3.1 Mash tun adjuncts<br />
Mash tun adjuncts fall into three classes, those that can be mixed into the grist without<br />
pre-cooking, such as wheat flours, those that are pre-cooked before mashing begins (e.g.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
flaked maize, torrefied wheat) <strong>and</strong> those that are cooked in the brewery as part of the<br />
mashing programme, such as maize-, rice- <strong>and</strong> sorghum-grits (Tables 2.1, 2.2). The type<br />
of adjunct that astarch-rich material produces is largely determined by the gelatinization<br />
temperature of its starch (Table 2.3; Chapter 4). If the starch granules swell <strong>and</strong> lose their<br />
structure <strong>and</strong> become susceptible to rapid enzyme attack (i.e. gelatinize) at temperatures<br />
low enough for the malt enzymes to remain active, then that material (e.g. wheat flour)<br />
need not be pre-cooked. However, if the starch has ahigh gelatinization temperature (e.g.<br />
maize) the material must be cooked at ahigh temperature to gelatinize the starch (either<br />
by flaking or in acooker at the brewery site) before it is mixed with the main malt mash<br />
at atemperature at which the malt enzymes can act.<br />
Raw barley grain has been used as an adjunct after hammer-milling or other kinds of<br />
dry-milling or wet-milling. It is an advantage to wash the grain before use (Briggs et al.,<br />
1981;Wieg,1973,1987).Theviability ofthisgrainisirrelevant.Itcontains -amylase (a<br />
proportion of which is insoluble) <strong>and</strong> some other hydrolases, as well as proteins<br />
inhibitory to some -amylases, proteases <strong>and</strong> limit dextrinase. Mashes containing much<br />
raw barley often need to be supplemented with enzyme mixtures from microbes<br />
containing -amylase, protease <strong>and</strong> -glucanase to convert the starch, to provide<br />
sufficient amounts of FAN <strong>and</strong> to degrade the relatively large amounts of -glucans that<br />
are present. Coarsely ground grain gives poor extract recoveries, but finely ground grain,<br />
while giving higher yields of extract, causes problems, in particular even greater<br />
quantities of -glucans are extracted. Because of practical difficulties, including the need<br />
for prolonged temperature programmed mashes <strong>and</strong> problems with the lack of desired<br />
character <strong>and</strong> raw-grain flavours in the products, interest in `barley brewing' has<br />
declined.<br />
Wheat has been processed in various ways, but most wheat is used as flour. Wheat<br />
flour milling is aspecialized process that, by aseries of roller milling <strong>and</strong> sieving steps,<br />
can produce material that is nearly pure endosperm tissue. By removing the germs <strong>and</strong><br />
bran the starch percentage in the product is increased while the protein, ash <strong>and</strong> oil<br />
contents are reduced. <strong>Brewing</strong> flours are prepared from soft wheats, <strong>and</strong> often the<br />
nitrogen contents of the flours are high. By using air classification fractions can be<br />
obtained that are depleted in protein <strong>and</strong> enriched in starch <strong>and</strong> so yield higher extracts.<br />
For example, air classification of aflour containing 9.5% protein gave afraction with<br />
only 7% protein. The nitrogen-reduced material is used in brewing while the nitrogenenriched<br />
material is used in making biscuits. H<strong>and</strong>ling flour is not simple. Special<br />
hoppers, usually equipped with vibrating feeds, are needed to ensure that the flour flows<br />
<strong>and</strong> specialized conveying equipment (vibrating or pneumatic) is needed. Flour dust<br />
mixed with air can form explosive mixtures <strong>and</strong> so all the usual precautions must be<br />
taken.<br />
To minimize h<strong>and</strong>ling problems, flours with `clumps' of starch granules, cell walls<br />
<strong>and</strong> protein have been prepared, with particle diameters of about 100 m(rather than the<br />
more usual 17 35 m) but perhaps the most convenient preparations are those in which<br />
the flour particlesare`agglomerated', thatis,boundtogetherwith asoluble bindersothat<br />
the material produces little dust <strong>and</strong> is h<strong>and</strong>led in agranular form. In the mash the<br />
granules disintegrate releasing the flour. Wheat flour has ahigh extract content (340 lë/<br />
kg, on dry, Table 2.1), <strong>and</strong> its use favours haze stability <strong>and</strong> especially head formation<br />
<strong>and</strong> retention. Raw or pre-gelatinized rye or millets, used in relatively small amounts,<br />
support head retention better than wheat (Stowell, 1985). However, wheat flour retards<br />
wort separation both because pentosans increase the viscosity of the wort <strong>and</strong> because<br />
they <strong>and</strong> proteins form fine particles that block the mash bed. Lipid micelles may also<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 2.1 Typical analyses of some starch-rich adjuncts<br />
Adjunct Moisture<br />
Hot water extract<br />
TN TSN Bulk density<br />
% (1ëkg as is) (1ëkg d.b.) (% d.b.) (% d.b.) (% d.b.) (kg/l)<br />
Maize grits 12 301 342 90 1.5 ± 0.76<br />
Maize flakes 9 313 344±355 ± 1.5 0.04 0.46±0.66 a<br />
Starches, refined 10 347 380±390 102±105 < 0.1 Neglible 0.60±0.70<br />
Rice grits 11 316 355 93 1.0 ± 0.85<br />
Rice flakes 9 325 357 ± 0.85 ± 0.30<br />
Wheat flour 11 304 342 ± 1.5 0.36 0.51<br />
Wheat, torrefied or micronized 4±9 291 310±315 ± 1.6±2.0 0.15 0.55<br />
Wheat flakes 5±8 287 302 ± 1.8 0.12 0.37±0.40<br />
Wheat, raw 12 260 295 ± 1.6 ± 0.77<br />
Barley, torrefied 5±6 254 267 72 1.8±2.2 ± 0.37±0.40<br />
Barley, flaked<br />
Barley, raw<br />
9<br />
12<br />
253<br />
250<br />
278 ± 1.8 0.12 0.25±0.26<br />
b<br />
284 b<br />
± 1.8 Variable b<br />
0.65±0.66<br />
Analyses IoB (1993) except HWE (%), determined by the ASBC method (ASBC, 1992).<br />
a Depends on the degree to which they are crushed.<br />
b Depends on added enzymes.<br />
From Lloyd (1986, 1988a); Brookes <strong>and</strong> Philliskirk (1987); Briggs (1998).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 2.2 Analysis (ASBC) of various adjuncts<br />
Adjunct Moisture<br />
Extract<br />
Protein Fat/oil Fibre Ash pH<br />
Gelatinization<br />
temperature range<br />
(% as is) (% as is) (% on dry) (% as is) (% as is) (% as is) (% as is) (ëC) (ëF)<br />
Corn (maize) grits 9.1±12.5 78.0±83.2 87.7±92.8 8.5; 9.5 0.1±1.1 0.7 0.3±0.5 5.8 61.6±73.9 143±165<br />
Corn (maize) flakes<br />
Refined maize grits<br />
4.7±11.3 82.1±88.2 91.0±93.4 ± 0.31±0.54 ± ± ± ± ±<br />
(maize starch) 6.5±12.3 90.6±98.3 101.2±105.6 0.4 0.04 ± ± 5.0 61.5±73.9 143±165<br />
Rice grits 9.5±13.4 80.5±83.8 92.2±96.1 5.4; 7.5 0.2±1.1 0.3±0.6 0.5±0.8 6.4 61.1±77.8 a<br />
142±172 a<br />
Sorghum grits 10.8; 11.7 81.7; 81.3 91.4; 91.1 8.7; 10.4 0.5; 0.65 0.8 0.3±0.4 ± 67.2±78.9 153±174<br />
Wheat flour 11.5 80.1 90.7 11.4 0.7 ± 0.8 ± ± ±<br />
Wheat starch 11.1; 11.4 86.5; 95.2 105.2; 97.5 0.2 0.4 ± ± 5.7 51.7±63.9 125±147<br />
Torrefied wheat 4.9 74.4 78.2 12.2 1.0 ± ± 6.2 ± ±<br />
Torrefied barley 6.0 67.9 72.2 13.5 1.5 ± ± 5.9 ± ±<br />
a Variance in US rice types: 65±68 ëC (149±154.4 ëF); 71±74 ëC (c. 160±165 ëF) (long grain rice).<br />
Data of Canares <strong>and</strong> Sierra (1976); Coors (1976); Bradee (1977); Canales (1979); through Briggs (1998).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 2.3 Some reported gelatinization temperature ranges of starches (Briggs, 1998; Reichelt,<br />
1983; Various). Reported values often disagree, probably because different methods have been used<br />
to determine them (Bentley <strong>and</strong> Williams, 1996). The values reported in ëF are only approximate<br />
equivalents of the temperature in ëC<br />
Gelatinization temperature range<br />
Starch ëC ëF<br />
Maize (Corn)* 62±77 143±171<br />
Waxy Maize* 62±80 143±176<br />
Sorghum* 69±75 156±167<br />
Millets* 54±80 129±176<br />
Barley 60±62 140±144<br />
Barley, small granules 51±92 124±198<br />
large granules 60±65 140±149<br />
Barley Malt 64±67 147±153<br />
Wheat 52±66 126±151<br />
Rye 49±61 120±142<br />
Oats 52±64 126±147<br />
Rice* 61±82 142±180<br />
Rice, short grain* 65±68 149±155<br />
Rice, long grain* 71±74 160±165<br />
Potato 56±71 133-160<br />
Tapioca 63±80 145±176<br />
Arrow root (Maranta) 67±85 152±185<br />
* Starches or adjuncts made from these materials must always be cooked before mashing. The other materials<br />
may be converted better if first cooked.<br />
contribute to these filtration problems, as they do with the filtration problems<br />
encountered with wheat starch hydrolysates (Matser <strong>and</strong> Steeneken, 1998). Sometimes<br />
these problems can be reduced by adding microbial pentosan-degrading enzymes to the<br />
mash.<br />
Wheat flour is now used to about 5 10% malt replacement in some British breweries<br />
although, in the past, much higher replacement rates, of 25% or even 36%, were used<br />
(Briggs et al., 1981). Wheat flour is used directly in infusion mashes, but higher extracts<br />
may be obtained if the flour is pre-soaked or is pre-cooked. From time to time purified<br />
starches from wheat, potatoes, manioc <strong>and</strong> other sources are used in mashing, depending<br />
on local economics. It might be expected that wheat starch would be fully converted in<br />
the mash, but it has been reported that better extract recovery occurs if the material is<br />
cooked at 96ëC (c. 205ëF), possibly because the small starch granules are gelatinized<br />
only at the high temperature. The material is not boiled to avoid frothing.<br />
Flours are produced, as by-products, during the manufacture of maize, rice <strong>and</strong><br />
sorghum grits. Like the grits these flours must be cooked before being mixed in with the<br />
malt mash. The extract yields of refined starches are high since, due to the uptake of the<br />
water of hydrolysis during conversion, 100 units (on dry) of starch give rise to<br />
103 105% units of dry sugars <strong>and</strong> dextrins, so extracts of 380 390 lë/kg (on dry), or<br />
102 105% (on dry) are obtained. Generally, purified maize starch is not used directly in<br />
breweries although when it is it is cooked. Most brewing sugars are prepared from maize<br />
starch.<br />
Pre-cooked adjunctsused inmashinginclude micronized<strong>and</strong>torrefied whole grains or<br />
flaked wheat or barley or flaked maize grits or flaked rice grits or flaked pearl barley<br />
(Tables2.1, 2.2). These materials are easily h<strong>and</strong>led<strong>and</strong>,becausethey have been cooked,<br />
they yield better extracts than the raw materials because their starches are gelatinized <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the hemicelluloses are partly degraded. They are easily broken up in malt mills. Whole<br />
grains of wheat or barley, graded to remove thin grains <strong>and</strong> with adjusted moisture<br />
contents, are cooked in hot air at 220 260ëC (428 500ëF). During cooling the softened<br />
material becomes firm <strong>and</strong> has amoisture content of about 4%. Torrefied barley may<br />
have an extract (on dry) of 267 lë/kg or 72%, while for torrefied wheat the values may be<br />
310 315 lë/kg or 78 80%. It is advantageous to use grains with low nitrogen contents,<br />
since a1% increase in nitrogen content (6.25% crude protein) will reduce the extract<br />
yield by 5%. Micronized cereal grains have similar properties. These are prepared by<br />
cooking in athin layer, moving beneath gas flame heated ceramic tiles which give out<br />
radiant heat (infra-red radiation). The grain, which should be carefully conditioned <strong>and</strong><br />
not heated for too long aperiod to obtain the best quality product, reaches atemperature<br />
of about 140ëC (284ëF) (Brookes <strong>and</strong> Philliskirk, 1987; South, 1992). Micronized grains<br />
may be cooled <strong>and</strong> mixed with the malt before it is milled. While it is still hot,<br />
micronized grain may be rolled to form flakes, which do not need to be dried.<br />
The older process, for flaking whole grains, or pearl barley or maize- or rice-grits,<br />
began by adjusting the moisture content of the material then, after a period of<br />
conditioning, cooking at 90 100ëC (194 212ëF), flaking it by passing it between hot<br />
rollers. Theflakes were dried inastream ofhot air,before cooling. At present (2004) it is<br />
not economic to use flaked rice but it, like flaked maize, is awell-liked adjunct that gives<br />
up its extract easily, in good yield, even in simple infusion mashes (Table 2.1). Flaked<br />
barley <strong>and</strong>, to an even greater extent, flaked pearl barley (grains from which the husk <strong>and</strong><br />
surface layers have been removed by abrasive milling; Briggs, 1978) give problems in<br />
brewing largely because they contain comparatively large amounts of -glucan. Flaked<br />
barley has been prepared sprayed with asolution of bacterial enzymes containing -<br />
amylase, -glucanase <strong>and</strong> probably protease. The product had an appreciable cold water<br />
extract <strong>and</strong> did not give rise to highly viscous worts or any of the other problems<br />
associated with -glucans. In the past flaked oats were used in making some stouts. They<br />
were described as being greyish, with low extracts of 252 282 lë/kg, <strong>and</strong> were rich in<br />
husk, protein <strong>and</strong> in oil that could readily become rancid. Experimentally it has been<br />
shown that milled, cooked <strong>and</strong> extruded cereals are convenient adjuncts (Briggs et al.,<br />
1986; Dale et al., 1989; Laws et al., 1986). These preparations seem to be used only in<br />
the preparation of some African beers.<br />
Grits are preparations of nearly pure starchy endosperm in which the starch granules<br />
are invested with protein <strong>and</strong> are enclosed with cell walls (Johnson, 1991). For brewing<br />
purposes these are prepared from maize (`corn'), rice or sorghum(Tables 2.1, 2.2). These<br />
gritsmustbecookedbeforebeingmixedwiththemainmaltmash.Thehightemperatures<br />
used (up to boiling or, when processed under pressure, even over 100ëC; 212ëF) disrupt<br />
the cellular structure of the grits <strong>and</strong> gelatinize the starch. The -amylase included in the<br />
mixture liquefies some of the starch, reducing the viscosity of the mixture <strong>and</strong> preventing<br />
retrogradation. The -amylase may be bacterial in origin or it may be from asmall<br />
amount of highly enzymic, ground malt. <strong>Brewing</strong> rice is usually aby-product of grain<br />
beingprepared forhumanconsumption.Thismaterialhasbecome tooexpensivetousein<br />
manyareas,butitisstillusedinAsia.Preferredricegritsarelessthan2mm(0.079in.)in<br />
diameter, have moisture contents of about 13%, <strong>and</strong> extracts of 88 90 or even 95% (on<br />
dry). Typically they contain 5 8% protein <strong>and</strong> 0.2 0.4% oil <strong>and</strong> about 0.9% ash. The<br />
flavour imparted to beers by rice are described as neutral, `dry', `light' <strong>and</strong> `clean'.<br />
Different grades of rice behave very differently in mashing, so that wort separation<br />
timesmayvarybyfactorsof2or3<strong>and</strong>thegelatinizationtemperaturesofthestarchesvary<br />
widely (Table 2.3). When rice grits are slurried in water <strong>and</strong> are progressively heated it is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
found that the gel point of each sample (the temperature at which the viscosity suddenly<br />
increases) is related to the difficulties of using the material in the brewhouse (Teng et al.,<br />
1983). It is advantageous to use athermostable bacterial -amylase when cooking rice. It<br />
seems that all rice grits should be heated to boiling. Rice grits, like maize grits, may be<br />
cooked <strong>and</strong> flaked. Flaked preparations are used in brewing without the need for acereal<br />
cooker.Typicalanalysesofmaizegrits,whicharepreparedfromyellowdentcornbyadry<br />
milling process, are: moisture 13%; protein 7 9%; fat, 0.7 1%; ash, 0.5 0.7%; extract<br />
about 90% (on dry). Usually particles have diameters of 0.3 1.5mm, 0.012 0.059 in.<br />
(Tables 2.1, 2.2; Johnson, 1991). These grits impart afuller flavour to beer, compared to<br />
rice grits. In some areas, notably Africa <strong>and</strong> Mexico, sorghum grits are used. In quality<br />
theycloselyresemblemaizegrits,givingextractsof91%(ondry),withmoisturecontents<br />
of 11 12%. When first used sorghum grits gave unpleasant flavours to beers, but<br />
improved milling techniques, processing large yellow or white, low tannin grains, now<br />
produce fully acceptable materials. Pearl barley is analogous to grits, being almost pure<br />
endosperm tissue. It does not need to be cooked but it is now little used in brewing. Grits<br />
can be prepared from several millets, but this is probably not done commercially.<br />
2.3.2 Copper adjuncts<br />
Copper adjuncts come in two categories (Table 2.4). First, wort extenders, which add<br />
essentially only carbohydrates (such as sucrose, invert sugar <strong>and</strong> hydrolyzed starch<br />
syrups) <strong>and</strong> wort replacements such as malt extracts <strong>and</strong> syrups made from hydrolyzed<br />
cereals. These materials add carbohydrates <strong>and</strong> acomplex mixture of other substances to<br />
the process stream. The formulae of sugars are given in Chapter 4.<br />
Sucrose (`sugar'), derived from either sugar cane or sugar beet, is awell-liked copper<br />
adjunct, used either as asolid or in solution <strong>and</strong> either as the disaccharide sucrose ( -Dglucopyranosyl-(1,2)-<br />
-D-fructofuranose) or as the hydrolysis product, the equimolecular<br />
mixture of glucose <strong>and</strong> fructose, `invert sugar', so called because as the sucrose is<br />
hydrolysed the optical rotation of the solution decreases <strong>and</strong> becomes negative <strong>and</strong><br />
`inverted'. At present, with the exception of Australia, sucrose-based materials are little<br />
used because they are costly. Beet sugar must be used pure, because the impurities have<br />
unpleasant flavours. While pure cane sugar is perfectly acceptable, partially purified<br />
preparations have been preferred because of their luscious flavours. These preparations<br />
may contain small quantities (perhaps 5%) of unfermentable di- <strong>and</strong> tri-saccharides<br />
(Table 2.4). Sucrose is extremely soluble (see Appendix), solutions containing over 63%<br />
solids being attainable. However, concentrated solutions of pure sugars are liable to<br />
crystallize. Solutions of invert sugar containing 83% solids can be prepared. Some<br />
brewer's preparations contain both sucrose <strong>and</strong> invert sugar. Yeasts ferment these sugars<br />
easily so, as the sugars dissolve completely in the wort, extract recovery is 100% <strong>and</strong>,<br />
with the pure preparations, the added sugars are 100% fermentable. The sugars may be<br />
provided in solution or as solids. A sugar syrup may give an extract of 258 lë/kg (fresh<br />
wt.), have a specific gravity of about 1.33 <strong>and</strong> a colour of 3 12 ëEBC. Nitrogen contents<br />
(e.g. 0.01%) are negligible. An invert sugar preparation may have an extract of 318 lë/kg<br />
(fresh wt.; Table 2.4). To prevent crystallization <strong>and</strong> to reduce the viscosity, so improving<br />
h<strong>and</strong>ling characteristics, these sucrose or invert sugar syrups are h<strong>and</strong>led <strong>and</strong> stored<br />
warm, at 40 50 ëC (104 122 ëF). They cannot be stored for long periods, <strong>and</strong> so must be<br />
delivered shortly before use.<br />
Sugar adjuncts used in small amounts include lactose (from whey; a sweet, nonfermentable<br />
sugar), honey <strong>and</strong> maple syrup (Wainwright, 2003). Many sugar preparations<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 2.4 Typical analyses of sugar-rich brewing adjuncts, priming sugars <strong>and</strong> caramels<br />
Preparation Hot water Total nitrogen Colour Fermenability Specific gravity<br />
extract (f.wt.) (%) 10% w/v solution solids (20 ëC)<br />
(f.wt.) (EBC units) (%)<br />
(1ë/kg) a<br />
Sugar preparations<br />
Raw cane sugar syrup 258 0.01 3 95+ 1.33<br />
Invert sugar liquid 318 0.01 3±12 95+ 1.43<br />
Sugar primings 310 0.01 3±12 95+ 1.42<br />
Refined maize-starch hydrolysates<br />
<strong>Brewing</strong> syrup 310 0.02 Colourless or 77±78 1.42<br />
adjusted b<br />
Confectioners' glucose 318 < 0.01 Colourless 30±50 1.43<br />
Solid brewing sugar 310 0.02 Colourless or 86±87 ±<br />
adjusted b<br />
Glucose chips 318 0.01 20±50 82 ±<br />
Other materials<br />
Grain-based syrup 302 0.4±0.8 4 65±70 1.40<br />
Malt extract 302 0.65±1.3 4 70 1.40<br />
Caramel, 46,000 liquid 242 ± 4600 ± 1.29<br />
Caramel, 32,000 liquid 284 ± 3200 ± 1.36<br />
a<br />
Dry, solid sugar preparations, e.g., sucrose, have values of 382±386 1ë/kg.<br />
b<br />
FAN values of 0.01±1.15%. The colour may be adjusted to specification by the addition of other sugar-based products.<br />
Analyses IoB.<br />
After Lloyd (1986, 1988a); through Briggs (1998).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
are made from `refined grits', refined maize starch (corn flour). This is prepared by a<br />
continuous wet milling process. Maize grain is soaked in a solution of sulphur dioxide<br />
<strong>and</strong> is then broken up. The oil-rich germs <strong>and</strong> hulls are separated <strong>and</strong> the remaining<br />
endosperm tissue is milled <strong>and</strong> the gluten <strong>and</strong> starch granules are separated. The starch is<br />
recovered as a 35 40% suspension in water. This material may be dried. The powdery<br />
product is dusty <strong>and</strong> must be h<strong>and</strong>led with all the precautions used with flours.<br />
Sometimes starch is added to the cereal cooker with grits, but it is usually converted into<br />
solutions of hydrolysis products by specialist manufacturers. A sample of maize starch<br />
had an amylose:amylopectin ratio of 28:72, a moisture content of 11 12%, a crude<br />
protein content of 0.35% <strong>and</strong> a fat content of 0.04 to 0.5%. The dry component of this<br />
material was mainly polysaccharide. Most maize starch is used in brewing, after<br />
hydrolysis, as syrupy copper (kettle) adjuncts. The starch is treated in two stages. In the<br />
first stage it is cooked to disrupt the granules <strong>and</strong> the material is treated with a mineral<br />
acid or a thermostable bacterial -amylase (stabilized by additions of calcium salts) to<br />
`liquefy' the polysaccharide, degrading it to a mixture of dextrins, oligosaccharides <strong>and</strong><br />
sugars. In part the high cooking temperature is needed to disrupt amylose-lipid<br />
complexes, making the polysaccharide more easily degraded. The liquefied mixture<br />
flows <strong>and</strong> has a comparatively low viscosity, in contrast to cooked, but not liquefied,<br />
starch which is very viscous <strong>and</strong> sets to a gel on cooling. The liquefied material is partly<br />
purified by treatment with active charcoal <strong>and</strong>/or ion exchange resins to remove lipids<br />
<strong>and</strong> ionic substances. If mineral acid was used then this must be neutralized if the next<br />
process is to be enzyme-catalysed. In the second stage the liquefied material is<br />
saccharified to produce the mixture of carbohydrates finally required. Saccharification<br />
may be carried out with mineral acid or, after adjustment of the pH, with one or more<br />
enzymes.<br />
A very wide range of products, varying in salt content, sugar spectra <strong>and</strong><br />
fermentability, are available. The materials may be classed as acid/acid, acid/enzyme<br />
or enzyme/enzyme products. Those prepared using acid hydrolysis may have high salt<br />
contents <strong>and</strong>, because of side reactions occurring during hydrolysis, may contain<br />
oligosaccharides containing unusual inter-sugar linkages, <strong>and</strong> may be coloured <strong>and</strong> have<br />
characteristic flavours. Thus acid/acid hydrolysis can yield confectioner's `chip sugar',<br />
which is rich in glucose <strong>and</strong> with colour in the range 200 500 ëEBC. Acid/enzyme <strong>and</strong><br />
enzyme/enzyme products may be produced with little colour <strong>and</strong> with closely controlled<br />
compositions. They may be dried <strong>and</strong> delivered as solids or in solution as liquid syrups.<br />
Generally, like the sucrose-based syrups, these syrups are kept warm (at 50 ëC; 122 ëF) or<br />
above) to prevent crystallization <strong>and</strong> the separation of solids from the mix <strong>and</strong> to reduce<br />
the viscosity. They may be delivered <strong>and</strong> stored at 60 70 ëC (140 158 ëF). The surfaces<br />
of the stored materials are often ventilated with sterile air to remove water vapour which<br />
otherwise might condense <strong>and</strong> drip back onto the surface of the syrup, so locally diluting<br />
it <strong>and</strong> allowing the proliferation of microbes, notably osmophilic yeasts. The headspace<br />
may be filled with nitrogen or be illuminated with sterilizing, ultraviolet light.<br />
Often syrups contain sulphur dioxide as a preservative (2 40 mg/l), <strong>and</strong> brewers<br />
specify an upper concentration. Syrups are described as having reducing dextrose<br />
equivalent (DE) values. However, as different mixtures of sugars <strong>and</strong> dextrins can have<br />
the same DE values, these are of limited use to brewers. Preparations can be obtained<br />
with fermentabilities ranging from 30 95%, but usually values are about 75 85%.<br />
These syrups can be used to adjust the final fermentability of wort. However, the<br />
fermentability of a syrup is not a sufficient characterization, the spectrum of sugars<br />
present is also significant. Thus the fermentable material may be rich in glucose, or be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
nearlyentirelyglucose.Thismaybeundesirablesinceyeastsinwortsrichinglucosemay<br />
not be able to adapt to metabolize maltose <strong>and</strong> maltotriose, leading to slow or `hanging'<br />
fermentations. Glucose-rich syrups are usually made with the enzyme amyloglucosidase,<br />
sometimes mixed with adebranching enzyme to accelerate the hydrolysis of the starch<br />
<strong>and</strong> dextrin -1,6-linkages. This problem does not arise if most of the fermentable<br />
carbohydrate is maltose. Maltose-rich syrups are made by incubating liquefied starch<br />
with a -amylase (a plant enzyme or the enzyme derived from Bacillus polymixa) <strong>and</strong> a<br />
debranching enzyme such as pullulanase.<br />
Starch-derived syrups <strong>and</strong> malt extracts <strong>and</strong> syrups prepared from cereal grains are<br />
introduced into the wort during the hop-boil, as are solid sugars (Chapter 10). All these<br />
materialsmustbedissolved<strong>and</strong>fullydispersed.Ifthisisnotachieved<strong>and</strong>materialsettles<br />
in the copper, the sugars can burn on to the heating surfaces with the creation of heating<br />
<strong>and</strong> cleaning problems, aloss of extract <strong>and</strong> perhaps the generation of unwanted flavours<br />
<strong>and</strong> colours. As prepared these syrups are very pale but, if required, makers may add<br />
caramel to give aspecified colour.<br />
Other copper adjuncts are malt extracts or syrups obtained by hydrolysing cereal<br />
grains (Briggs, 1978, 1998; Tables 2.4 <strong>and</strong> 2.5). These materials contain both<br />
carbohydrates <strong>and</strong> a complex mixture of substances including nitrogenous materials,<br />
minerals <strong>and</strong> yeast growth factors. Additions of these materials to the wort are equivalent<br />
to adding concentrated wort to the beer production stream. Malt extracts are made by<br />
grinding the malt, mashing it, with or without mash tun adjuncts <strong>and</strong> supplementary<br />
enzymes, <strong>and</strong> separating the wort, then concentrating it using triple effect vacuum<br />
evaporators. Many types of material can be produced depending on the grist, the mashing<br />
programme employed <strong>and</strong> the evaporation conditions used. By mashing enzyme-rich<br />
malts at low temperatures <strong>and</strong> concentrating the worts at low temperatures, enzyme-rich<br />
malt extracts may be obtained. At the other extreme, by heating the wort strongly,<br />
sometimes at a reduced pH, before concentration a product lacking enzymes can be<br />
prepared. Extracts can contain 75 82% solids (SG values 1400 1450), the more<br />
concentrated materials being used in the tropics. To keep the preparations liquid they<br />
need to be kept warm (e.g. 50 ëC, 122 ëF). At this temperature the material will slowly<br />
continue to darken <strong>and</strong> its other characteristics will change, so it should be used promptly.<br />
A representative extract is 302 lë/kg (fresh wt.). Colours may range from 3 520 ëEBC,<br />
have varied enzyme contents (DP 0 400 ëL), <strong>and</strong> fermentabilities in the range 56 93%.<br />
Some of the mash grists contain large proportions of raw cereal or cereal adjuncts, <strong>and</strong><br />
to obtain adequate extracts the mashes may be supplemented with microbial enzymes <strong>and</strong><br />
long, rising temperature programmes may be used. These products are best termed cereal<br />
syrups. A distinct product was `liquid malt'. This was made by mashing green barley<br />
malt, so eliminating the cost of kilning. The wort was concentrated in the usual way <strong>and</strong><br />
the unwanted flavour components were evaporated during the concentration stage.<br />
According to German law this material is not an adjunct <strong>and</strong> so, like conventional malt<br />
extracts, its use is permitted. Potentially such syrups are highly fermentable, can be<br />
enzyme rich <strong>and</strong> low in proanthocyanidins (anthocyanogens), <strong>and</strong> so their use favours<br />
haze stability in beers. Malt extracts <strong>and</strong> cereal syrups are used less by large-scale<br />
breweries than was once the case. In contrast syrups made by hydrolysing starches are<br />
widely used. While malt extracts were once added to supplement the enzyme contents of<br />
mashes this highly uneconomic <strong>practice</strong> has long been discontinued, at least in largescale<br />
brewing. However, 3.7 volumes of malt extract give about the same amount of<br />
extract as 10 volumes of malt, making it a very compact source of extract, <strong>and</strong> it has been<br />
the <strong>practice</strong> to send malt extract (pre-hopped or not) to be fermented to make beer at<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 2.5 The carbohydrate compositions (%) of two worts <strong>and</strong> several syrups prepared from starches (after Wainwright, 2003)<br />
Infusion Decoction Low fermentable 63 DE High maltose Very high High dextrose<br />
mash wort mash wort syrup* syrup* syrup* maltose syrup* (glucose) syrup*<br />
Glucose + Fructose 10 11 4 38 2 3 94<br />
Sucrose 5 2 0 0 0 0 0<br />
Maltose 45 52 10 33 55 68 3.5<br />
Maltotriose 15 12 12 6 16 18.5 0<br />
Dextrins y 25 23 74 23 27 10.5 2.5<br />
Fermentability (%) z 72 74.8 23.6 75.6 69.8 85.8 97.5<br />
* Starch hydrolysates do not contain fructose or sucrose.<br />
y Dextrins are not fermentable.<br />
z Calculated fermentability.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
emote locations or on ships. In addition, hopped or unhopped malt extracts are used by<br />
many home brewers<strong>and</strong> small-scale or`micro' brewers toavoid the inconvenience of the<br />
mashing <strong>and</strong> wort separation operations. Copper adjuncts effectively increase the<br />
production capacity of abrewery. They are convenient for preparing high-gravity worts<br />
<strong>and</strong> for adjusting wort fermentability. Most add insignificant amounts of nitrogenous<br />
substances or polyphenols, or flavours or colours to wort. The approved sets of analytical<br />
methods specify ways of evaluating copper adjuncts. In contrast to malts <strong>and</strong> mash tun<br />
adjuncts all the potential extract of acopper extract is recovered in the wort provided that<br />
it is completely dissolved.<br />
2.4 Priming sugars, caramels, malt colourants <strong>and</strong> Farbebier<br />
The materials described in this section are not regarded as adjuncts. However, they all<br />
add extract to the wort or the product <strong>and</strong> so they are considered here. Priming sugars are<br />
addedtobeersthataretobecask-orbottle-conditioned.Theobjectistoprovidetheyeast<br />
with asupply of easily fermented sugars that can indirectly supply the carbon dioxide<br />
needed to carbonate the beer, <strong>and</strong> `bring the beer into condition'. Since the sugars are<br />
mostly fermented their nature is not important; sucrose, invert sugar <strong>and</strong> glucose- or<br />
maltose-rich syrups will serve. However, if the preparation contains aproportion of<br />
unfermentable material this will remain in the beer <strong>and</strong> may alter its character. To utilize<br />
some of the residual dextrins present in beer, enzymes have been added to catalyse the<br />
hydrolysis of aproportion into fermentable sugars, aprocedure which removes the need<br />
for priming sugars. Various enzymes have been used for this purpose. Amyloglucosidase<br />
was unsatisfactory since it is too stable <strong>and</strong> so is not reliably destroyed by the<br />
temperatures reached during pasteurization. Consequently the enzyme continues to act<br />
<strong>and</strong> sweeten the beer when its activity is no longer required. Less stable enzymes such as<br />
fungal -amylase, or pullulanase with -amylase have been more successful, but the<br />
problem of deciding when the correct degree of dextrin degradation has occurred, <strong>and</strong> so<br />
when the enzymes must be inactivated, still remains. Sugars may be added to some<br />
filtered <strong>and</strong> sterile beers to sweeten them. If this is done then sucrose or high-fructose<br />
preparations are probably to be preferred.<br />
Caramels are used to adjust colour by adding them to the wort or beer (Chapter 9).<br />
Caramels are made in different ways <strong>and</strong> not all types are suitable for brewing purposes<br />
(Comline, 1999). The class III, electropositive-ammonia caramels, the caramels used in<br />
beers, are made by heating sugars (usually high glucose syrups) with ammonia. Complex<br />
reactions occur <strong>and</strong> the product is a mixture of high molecular weight coloured<br />
substances <strong>and</strong> lower molecular weight substances which impart flavour <strong>and</strong> aroma. The<br />
preparations may have colours up to 35,000 ëEBC, contain 65 75% solids, 2.5 5%<br />
nitrogen <strong>and</strong> have pI values of 6.0 6.5. (A pI value of a substance is the pH at which it is<br />
50% ionized). By using ultrafiltration the coloured <strong>and</strong> flavoured components can be<br />
separated, permitting beer colour <strong>and</strong> flavour to be adjusted separately (Walker <strong>and</strong><br />
Westwood, 1991). The specifications of brewing caramels usually include values for<br />
colour, pH, extract content, <strong>and</strong> stability when dissolved in worts <strong>and</strong> beers.<br />
Sometimes the use of caramels is forbidden but it is permissible to use extracts from<br />
coloured malts. Crystal, chocolate or black malts (or roasted barley, where allowed) are<br />
extracted with hot water <strong>and</strong> the extracts are concentrated. Colours (of 10% solutions) of<br />
850 1,700 ëEBC may be obtained. It is not clear how widely these malt colourants are<br />
used. In Germany beer colour may be adjusted using Farbebier. This `colour beer' is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
produced by specialist manufacturers (Narziss, 1992; Kunze, 1996; Riese, 1997). A<br />
mixture of, say, 60% pale malt <strong>and</strong> 40% dark malt is mashed to give awort with avery<br />
high density (e.g. 18 20ëPlato, approx. SG 1074 1084). The extract is boiled <strong>and</strong><br />
fermentedinaspecialwaytogiveaproductwithacolourofabout8,000ëEBC.Itmaybe<br />
concentrated under vacuum. This material is undrinkable, but it is added to wort or beer<br />
to adjust the colour. Sometimes the material is treated with active charcoal to reduce<br />
bitter flavour.<br />
2.5 Supplementary enzymes<br />
Enzymes derived from sources other than malt may be used at various stages during<br />
brewing, provided that thisis allowed by local regulations (Bamforth,1986; Briggs et al.,<br />
1981; Byrne, 1991; Godfrey <strong>and</strong> Reichelt, 1983). Enzymes are also used in the<br />
production of some adjuncts (Section 2.3). These enzymes are mostly prepared using<br />
liquid suspension cultures of various microbes (bacteria <strong>and</strong> fungi), but a few are<br />
obtained from plants <strong>and</strong> at least one was obtained from animal sources. The<br />
preparations, which may be dry powders or solutions, must be approved for use in<br />
foodstuffs. They are not `pure', <strong>and</strong> will usually contain residual materials from the<br />
nutrient medium in which the microbes were cultured, other enzymes besides the one(s)<br />
specified, diluents, extenders or carriers, <strong>and</strong> preservatives. They should not contain<br />
viable microbes. The preparations available have a wide range of characteristics.<br />
Different suppliers describe their preparations in different ways so that it is difficult to<br />
make comparisons between them. The lack of st<strong>and</strong>ard analyses is a source of difficulties.<br />
The temperature <strong>and</strong> pH optima of enzymes are so influenced by incubation conditions,<br />
<strong>and</strong> the conditions used in different breweries <strong>and</strong> at different stages of brewing are so<br />
varied, that it is not possible to give useful values. Consequently the effectiveness of the<br />
addition of an enzyme preparation must be determined by brewers under their particular<br />
processing conditions.<br />
The activities of `named' enzymes in preparations are st<strong>and</strong>ardized by suppliers.<br />
However, this is not true of other enzymes that may be present. The presence of these<br />
additional enzymes may be advantageous or harmful. For example, the presence of -<br />
glucanase in preparations of bacterial -amylase may be beneficial when added to a<br />
mash, particularly if undermodified malt or barley or oats adjuncts are used in the grist.<br />
On the other h<strong>and</strong>, while the presence of protease activity may be an advantage if more<br />
FAN is needed, it is most undesirable if it elevates the levels of soluble nitrogen too far<br />
<strong>and</strong>/or if the degradation of protein leads to a reduction in foam formation or stability.<br />
The presence of some `additional' enzymes can easily be detected (Albini et al., 1987).<br />
Enzyme preparations are not stable, so they should be stored cool <strong>and</strong> used fresh.<br />
Different enzymes in a mixture will usually have different half-lives, so the ratios of<br />
enzyme activities in a preparation will alter with storage time. This may generate<br />
problems. Many of the enzymes used in the manufacture of starch- <strong>and</strong> cereal-derived<br />
syrups may also be used in breweries. Usually enzymes, where used, may be added to the<br />
mash or the cooker, or they may be added to the wort or beer. Used intelligently they can<br />
improve extract recovery, wort collection rate, the rate of beer filtration <strong>and</strong> the length of<br />
filtration runs, wort fermentability, <strong>and</strong> the resistance of the beer to haze formation.<br />
Added enzymes can minimize the presence of residual starch or gums in the wort. Other<br />
uses are indicated later. The enzymes of most interest in brewing are those which catalyse<br />
the hydrolysis of starch <strong>and</strong> dextrins, those which attack hemicelluloses <strong>and</strong> gums (both<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
-glucans <strong>and</strong> pentosans), <strong>and</strong> those which degrade proteins. However, other enzymes<br />
may be of interest. While some brewers may use added enzymes on a routine basis others<br />
use them only to combat unforeseen production problems.<br />
-Amylases used in brewing are from different sources <strong>and</strong>, because of their different<br />
properties, they are suited to different purposes. They are all stabilized by elevated levels<br />
of calcium ions <strong>and</strong> by their substrates, starch <strong>and</strong> dextrins. They are all endo-acting<br />
enzymes, that is they catalyse the hydrolysis of the -(1, 4)-links within the dextrin,<br />
amylose <strong>and</strong> amylopectin chains. However, the range of hydrolysis products differs<br />
significantly, <strong>and</strong> the enzymes differ in thermal stabilities to a remarkable extent. Fungal<br />
enzymes (usually from Aspergillus spp.) have pH optima in the range 5.0 6.5, <strong>and</strong><br />
temperature optima of around 60 65 ëC (140 149 ëF), or 55 ëC (131 ëF). Despite these<br />
low values these preparations have been added to mashes where the complex mixture of<br />
`extra' enzymes (which commonly include hemicellulases <strong>and</strong> proteases) may be of<br />
value. This type of enzyme, which is inactivated by pasteurization, has been added to<br />
beers to hydrolyse dextrins <strong>and</strong> so obviate the need for priming sugars. It produces<br />
appreciable amounts of maltose among its products.<br />
Several types of bacterial -amylase are in use. The enzyme from Bacillus subtilis has<br />
a pH optimum between 6.0 <strong>and</strong> 7.5, but it is usefully active at mash pH values, 5 6. The<br />
temperature optimum is around 65 70 ëC (149 158 ëF), but is strongly dependent on the<br />
presence of starch, which stabilizes it. The enzyme may act briefly at temperatures up to<br />
80 ëC (176 ëF). Usually preparations of this enzyme, like those other bacterial enzymes,<br />
contain protease <strong>and</strong> -glucanase activities. While the alkaline protease may have little<br />
action under mashing conditions the neutral protease does. The enzyme from another<br />
bacterium, Bacillus subtilis, var. amyloliquefaciens is appreciably more heat stable. This<br />
-amylase has a reported pH optimum at 5.7 5.9 (at 40 ëC; 104 ëF). Although its<br />
temperature optimum is about 70 ëC (158 ëF) this enzyme is able to liquefy a 35 40%<br />
starch slurry at 85 90 ëC (185 194 ëF), <strong>and</strong> so it is useful for liquefying the starch when<br />
adjuncts are cooked, since it is so much more stable than the malt enzyme. In contrast the<br />
-amylase from Bacillus licheniformis is too heat stable for some brewing purposes. This<br />
enzyme, which has a wide pH optimum around 6, has a temperature optimum at 90 ëC<br />
(184 ëF) at high calcium ion concentrations. It can act briefly at 115 ëC (239 ëF), <strong>and</strong> it is<br />
not reliably destroyed by boiling unless the solution is slightly acid <strong>and</strong> the calcium <strong>and</strong><br />
starch concentrations are low. These conditions can be met when the enzyme is used to<br />
liquefy starch during the manufacture of sugars <strong>and</strong> syrups, but cannot be reliably<br />
achieved in brewing.<br />
Debranching enzymes are used in the manufacture of copper adjuncts, <strong>and</strong> they have<br />
been investigated for use in the brewhouse. Two types of enzyme have been investigated.<br />
Isoamylase is able to hydrolyse the -(1,6)-links in amylopectin but not in dextrins. This<br />
enzyme seems not to be of value in brewing. However, pullulanase, an enzyme produced<br />
by the bacterium Klebsiella pneumoniae (Aerobacter aerogenes), hydrolyses -(1,6)links<br />
in both amylopectin <strong>and</strong> in dextrins, including limit dextrins. The enzyme is<br />
thermolabile, <strong>and</strong> is used at 45 55 ëC (113 131 ëF), when saccharifying dextrins with<br />
amyloglucosidase or -amylase in making glucose- or maltose-rich syrups respectively.<br />
The enzyme has been added to cooled mashes in experimental brewing, <strong>and</strong> it has been<br />
used, together with -amylase, to replace priming sugars in beer. As it is readily<br />
inactivated by heat this process can be stopped by pasteurizing the beer. The pH optimum<br />
has been given as 5.5 6.0, but the enzyme has been used at values as low as 4.<br />
-Amylases may be obtained from plants or particular bacteria. Enzymes from cereals<br />
(including flours), soya beans <strong>and</strong> sweet potatoes have been used to saccharify dextrins,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
with or without the addition of other hydrolases. The pH optima are about 5.3, but the<br />
useful pH range is about 4.5 7.0. These enzymes attack the penultimate -(1,4)-links in<br />
starch chains, releasing the disaccharide maltose. They are readily denatured by heat, <strong>and</strong><br />
have temperature optima around 55ëC (131ëF). These enzymes have been added to<br />
mashes to increase the wort fermentability, <strong>and</strong> they have been added to wort for the<br />
same purpose <strong>and</strong> to beers to replace priming sugars. These enzyme preparations often<br />
contain -glucosidase (which is generally ignored) <strong>and</strong> may contain the unwanted<br />
enzyme lipoxygenase (LOX) as well as other enzymes. The bacteria Bacillus polymixa<br />
<strong>and</strong> Bacillus cereus, var. mycoides, produce both pullulanase <strong>and</strong> -amylase which,<br />
acting together, have been used when making high maltose syrups.<br />
Amyloglucosidase (syn. glucoamylase; AG; AMG) is prepared from several different<br />
fungi (e.g. Aspergillus spp., Rhizopus spp.). Some preparations also contain -amylase<br />
<strong>and</strong>/or transglucosidase. The latter is undesirable as it catalyses the formation, by<br />
transglucosylation, of unwanted <strong>and</strong> unfermentable oligosaccharides such as isomaltose<br />
<strong>and</strong> panose. Amyloglucosidase attacks the non-reducing ends of starch chains <strong>and</strong><br />
dextrins releasing glucose. Its attack on -(1,4)-links is comparatively rapid relative to<br />
the attack on -(1,6)-links, so the conversion of starch into glucose by this enzyme is<br />
accelerated by the addition of pullulanase. The optimal pH range is 4.0 5.5, <strong>and</strong> the<br />
enzyme will act for extended periods at 60 65ëC (140 149ëF). It has been added to<br />
mashes (particularly mashes containing large proportions of adjuncts) to increase the<br />
fermentability of the wort. It is regularly used in the production of glucose <strong>and</strong> has been<br />
added to beer to replace priming sugars. It is no longer used for this, being replaced by<br />
more thermolabile enzymes.<br />
There is aproposal to add aglycosyl transferase to mashes to increase the levels of<br />
unfermentable isomaltooligosaccharides in the wort to produce abeer with areduced<br />
alcohol content but with a full body. In contrast, the same enzyme added to cool,<br />
fermenting wort increases the fermentability <strong>and</strong> hence the final alcohol content<br />
(Robinson et al., 2001).<br />
Whenundermodified orinhomogeneousbarleymaltsareusedorwhenbarley(oroats)<br />
mashtunadjunctsareemployed,problemscan ariseinthebrewery<strong>and</strong>theseareoften,at<br />
least partly, due to residual, high molecular weight -glucans. Similarly, when problems<br />
arise from the use of wheat, rye or triticale adjuncts or wheat malt the problems are often<br />
attributed to pentosans. The problems include slow wort separation, slow beer filtration<br />
<strong>and</strong> short filter runs <strong>and</strong> sometimes the separation of hazes <strong>and</strong> gelatinous precipitates in<br />
the beer. The enzymes used to degrade -glucans may be divided into -glucanases <strong>and</strong><br />
cellulases. Sometimes these preparations contain complex mixtures of enzymes. Because<br />
the structures of pentosans are complex (Chapter 4) mixtures of enzymes may be needed<br />
to obtain substantial degradation of these materials.<br />
The -glucanase of Bacillus subtilis is a well characterized enzyme, with an optimal<br />
pH range of 6.0 7.5 <strong>and</strong> temperature range of 50 60 ëC (122 140 ëF). In temperatureprogrammed<br />
mashes it acts best at about 50 ëC (122 ëF). However, the enzyme will act in<br />
brewery mashes, at least briefly, at about pH 5.3, at temperatures up to 75 ëC (167 ëF).<br />
This enzyme is specific in that it attacks only mixed-linked -(1,3;1,4)-glucans. It has<br />
been used in mashes made with barley adjuncts, <strong>and</strong> it is usually accompanied by -<br />
amylase <strong>and</strong> two proteases. Fungal -glucanase preparations (e.g. from Aspergillus spp.)<br />
have varied properties, but usually have inconveniently low temperature optima<br />
(45 60 ëC; 113 140 ëF) for mashing but have convenient pH optima in the range<br />
3.5 6.0. They probably contain a complex mixture of hydrolases, <strong>and</strong> are not clearly<br />
distinguished from the cellulases.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Apreparation from Humicola insolens is active at degrading -glucans at up to 75ëC<br />
(167ëF). Cellulases used in brewing include those from Trichoderma spp. (T. reesei; T.<br />
viride), with temperature optima of 50 55ëC (122 131ëF) <strong>and</strong> pH optima in the range<br />
3.5 5.5. Such preparations are useful in temperature programmed brewery mashes. They<br />
contain mixtures of enzymes, including amylases <strong>and</strong> pentosanases. Cellulase preparationsfromPenicilliumfuniculosumhaveactivityinthepHrange4.3<br />
5.0,<strong>and</strong>functionat<br />
temperatures of 65ëC (149ëF). Preparations from P. emersonii are more heat stable, with<br />
an optimal temperature of 80ëC (176ëF) <strong>and</strong> auseful optimum pH range of 3.7 5.0. The<br />
enzyme mixture attacks not only mixed link barley -glucans but also holocellulosic<br />
material in barley, starch <strong>and</strong> pentosans. It is well suited for addition to mashes.<br />
Pentosanases need to be complex mixtures of enzymes <strong>and</strong> contain acetyl esterase,<br />
feruloylesterase, -L-arabinofuranosidase, exo-(xylobiase) <strong>and</strong>endo-xylanase activities.<br />
Preparations usually contain starch-, cellulose- <strong>and</strong> -glucan-degrading activities.<br />
Preparations have been made from Disporotrichum, Trichoderma <strong>and</strong> Aspergillus spp.<br />
Usually these are used in temperature-programmed mashes, being active at about 50ëC<br />
(122ëF). They are particularly useful when wheat, rye <strong>and</strong> triticale adjuncts are used.<br />
Non-maltproteolytic enzymes areusedfor twopurposes inbrewing.First, byaddinga<br />
protease to the mash, the amount of nitrogenous yeast nutrients (FAN; formol-N) in the<br />
wortisincreased<strong>and</strong>,secondly,byaddingaproteasetobeer,polypeptidehazeprecursors<br />
are degraded. Pepsin, an animal protease with an acidic pH optimum, was added to beer<br />
as astabilizing agent, but this function is now carried out by thiol-dependent plant<br />
proteases, in particular papain, from the latex of the pawpaw (Carica papaya), bromelin<br />
from the pineapple (Ananas spp.) <strong>and</strong> ficin from figs (Ficus spp.). These enzymes differ<br />
alittle in their properties. All are destroyed by pasteurization <strong>and</strong>, while they degrade<br />
hazeprecursors,theyapparentlydonotdegradethedesirablefoam-formingpolypeptides.<br />
In contrast the bacterial proteases do destroy foam precursors. Protease activities are<br />
often present in fungal enzyme preparations, with pH optima in the range 3 6, <strong>and</strong><br />
temperature optima around 50ëC (122ëF). Probably the bacterial proteases are of most<br />
interest, since these have been added to mashes to increase the FAN levels. Bacillus<br />
subtilis, like some other Bacilli, produces proteases having neutral <strong>and</strong> alkaline pH<br />
optima. The mixture works at mash pH values, <strong>and</strong> has an optimal temperature of<br />
45 50ëC (113 140ëF). When mashes are made that are rich in raw barley <strong>and</strong> are<br />
supplemented with bacterial enzymes an extended st<strong>and</strong> is needed at 50ëC (122ëF) to<br />
obtain an adequate level of soluble nitrogen in the wort. Raw barley contains an inhibitor<br />
of bacterial neutral protease.<br />
Lipases, nucleases, phosphatases (including phytase), oxidases, transglycosylases, -<br />
<strong>and</strong> -glucosidases are enzymes of potential interest. It is generally considered that the<br />
presence of lipases (fat hydrolysing enzymes) <strong>and</strong> lipoxygenase is undesirable. It has<br />
been proposed that the addition of `tanninases' to wort or beer should hazeproof the beer.<br />
How, or if, such enzymes might act on barley <strong>and</strong> hop proanthocyanidins is not clear,<br />
since these are not hydrolysable tannins. The enzyme -acetolactate decarboxylase may<br />
be added to beer to break down its substrate to carbon dioxide <strong>and</strong> acetoin. Thus by<br />
destroying aprecursor of diacetyl the flavour stability of the beer is improved (Section<br />
12.10.2). Glucose oxidase has been added to beer to `scavenge' oxygen, which is utilized<br />
to convert glucose to gluconolactone <strong>and</strong> hydrogen peroxide. This latter compound is<br />
itself an oxidizing agent <strong>and</strong> its presence is probably undesirable. It is degraded by the<br />
enzyme catalase to oxygen <strong>and</strong> water. The oxygen (half the amount initially used) is<br />
again used by the glucose oxidase, <strong>and</strong> so the amount present is progressively reduced.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
2.6 References<br />
AASTRUP, S., BRANDT, J. <strong>and</strong> RIJS, P. (1991) Louvain <strong>Brewing</strong> Letters, 4 (1), 16.<br />
ALBINI, P. A., BRIGGS, D. E. <strong>and</strong> WADESON, A. (1987) J. Inst. <strong>Brewing</strong>, 93, 97.<br />
BAMFORTH, C. W. (1986) European Brewery Convention Monograph ± XI. E. B. C. ± Symposium on Wort<br />
Production, Maffliers. p. 149.<br />
BEMMENT, D. W. (1985) The Brewer, Dec., p. 457.<br />
BENTLEY, I. S. <strong>and</strong> WILLIAMS, E. C. (1996) in Industrial Enzymology (2nd edn, Godfrey, T. <strong>and</strong> West, S.,<br />
eds), Macmillan Press, London. pp. 339±57.<br />
BLANCHFLOWER, A. J. <strong>and</strong> BRIGGS, D. E. (1991) J. Sci. Food Agric. 56, 103, 117, 129.<br />
BOIVIN, P. <strong>and</strong> MALANDA, M. (1998) Proc. 25th Conv. Inst. <strong>Brewing</strong> (Asia Pacific Sect.), Perth. p. 30.<br />
BOURNE, D. T. <strong>and</strong> WHEELER, R. E. (1982) J. Inst. <strong>Brewing</strong>, 88, 324.<br />
BRIGGS, D. E. (1978) Barley. Chapman & Hall, London. 612 pp.<br />
BRIGGS, D. E. (1987) in Cereals in a European Context (Morton, I. D. ed.). Ellis Horwood, Chichester. p.<br />
119.<br />
BRIGGS, D. E. (1998) Malts & Malting. Blackie Academic & Professional, London. 796 pp.<br />
BRIGGS, D. E. (2002) J. Inst. <strong>Brewing</strong>, 108, (4), 395.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting & <strong>Brewing</strong> Science, Vol I, Malt<br />
& Sweet Wort (2nd edn). Chapman & Hall, London. 387 pp.<br />
BRIGGS, D. E., WADESON, A., STATHAM, R. <strong>and</strong> TAYLOR, J. F. (1986) J. Inst. <strong>Brewing</strong>, 92, 468.<br />
BRISSART, R., BRAÈ UNINGER, U., HAYDON, S., MORAND, R., PALMER, G., SAUVAGE, R. <strong>and</strong> SEWARD, B. (2000)<br />
European Brewery Convention Manual of Good Practice. Malting Technology. Fachverlag Hans<br />
Carl. NuÈrnberg. 224 pp.<br />
BROOKES, P. A. <strong>and</strong> PHILLISKIRK, G. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid. p. 337.<br />
BUCKEE, G. K. (1997) J. Inst. <strong>Brewing</strong>, 103, 115.<br />
BYRNE, H. (1991) Brew. Distill. Internat., 22 (11), 24.<br />
BYRNE, H. <strong>and</strong> LETTERS, R. (1992) Proc. 22nd Conv. Inst. <strong>Brewing</strong> (Australia & New Zeal<strong>and</strong> Sect.),<br />
Melbourne, p. 125.<br />
BYRNE, H., DONNELLY, M. F. <strong>and</strong> CARROLL, M. B. (1993) Proc. 4th Sci. Tech. Conf. Inst. <strong>Brewing</strong> (Central<br />
& Southern Africa Sect.), p. 13.<br />
CHANDRA, S., BOOER, C., PROUDLOVE, M. <strong>and</strong> JUPP, D. (1999) Proc. 27th Congr. Eur. Brew. Conv.,<br />
Cannes. p. 501.<br />
COMLINE, P. (1999) Brew. Distill. Internat. Sept., p. 14.<br />
COPESTAKE, C. (1998) The Brewer, Feb., p. 79.<br />
DALE, C. J., YOUNG, T. W. <strong>and</strong> MAKINDE, A. (1989) J. Inst. <strong>Brewing</strong>, 95, 157.<br />
DONHAUSER, S., WEIDENEDER, A. <strong>and</strong> GEIGER, E. (1991) Brauwelt Internat., 141, (4), 294.<br />
GIBSON, G. (1989) in Cereal Science & Technology (Palmer, G. H. ed.). p. 279. The University Press,<br />
Aberdeen.<br />
GODFREY, T. <strong>and</strong> REICHELT, J. (eds) (1983) Industrial Enzymology. The application of enzymes in<br />
industry. Macmillan, Basingstoke. 582pp.<br />
GRETENHART, K. E. (1997) MBAA Tech. Quart., 34 (2), 102.<br />
GROMUS, J. (1988) Brauwelt Internat., (II), 150.<br />
HAIKARA, A., ULJAS, H. <strong>and</strong> SUURNAÈ KKI, A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo. p. 163.<br />
HYDE, W. R. <strong>and</strong> BROOKES, P. A. (1978) J. Inst. <strong>Brewing</strong>, 84, 167.<br />
JOHNSON, L. A. (1991) in H<strong>and</strong>book of Cereal Science & Technology (Lorenz, K. J. <strong>and</strong> Kulp, K. eds).<br />
Marcel Dekker, New York. p. 55.<br />
KUNZE, W. (1996) Technology Malting <strong>and</strong> <strong>Brewing</strong>. (Internat. ed., Wainwright, T. transl.). VLB, Berlin.<br />
726 pp.<br />
LAITILA, A., SCHMEDDING, D., VAN GESTEL, M., VLEGELS, P. <strong>and</strong> HAIKARA, A. (1999) Proc. 27th Congr.<br />
Eur. Brew. Conv., Cannes, p. 559.<br />
LAWS, D. R. J., BAXTER, E. D. <strong>and</strong> CRESCENZI, A. M. (1986) European Brewery Convention Monograph- XI.<br />
E. B. C.-Symposium on Wort Production, Maffliers, p. 14.<br />
LETTERS, R. (1990) Louvain <strong>Brewing</strong> Letters, 4 (3/4) 12.<br />
LLOYD, W. J. W. (1986) J. Inst. <strong>Brewing</strong>, 92, 336.<br />
LLOYD, W. J. W. (1988a) Brewer, 74 (882), 147.<br />
LLOYD, W. J. W. (1988b) Brewers' Guard., 47, (5), 23.<br />
MARTIN, P. A. (1978) Brewers' Guard., Aug., p. 29.<br />
MATSER, A. M. <strong>and</strong> STEENEKEN, P. A. (1998) Cereal Chem., 75 (3), 189.<br />
MAULE, A. P. (1998) Ferment, 11 (1), 23.<br />
NARZISS, L. (1976) Die Bierbrauerei. Vol. I. Die Technologie der Malzbereitung. (6th edn). Ferdin<strong>and</strong><br />
Enke Verlag, Stuttgart. 382 pp.<br />
NARZISS, L. (1991) Brauwelt Internat., (4), 284.<br />
NARZISS, L. (1992) Die Bierbrauerei, Vol. II. Die Technologie der WuÈrzebereitung. (7th edn). Ferdin<strong>and</strong><br />
Enke Verlag, Stuttgart. 402 pp.<br />
REICHELT, J. R. (1983) in Industrial Enzymology. The application of enzymes in industry, (Godfrey, T. <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Reichelt, J. R. eds), Macmillan, Basingstoke.<br />
RIESE, J. C. (1997) MBAA Tech Quart., 34 (2), 91.<br />
ROBINSON, N., AMANO, H. <strong>and</strong> MIZUNO, A. (2001) Eur. Brew. Conv. Monograph no. 31. Symposium on<br />
flavour <strong>and</strong> flavour stability.<br />
SCHWARZ, P. B., JONES, B. L. <strong>and</strong> STEFFENSON, B. J. (2002) J. Amer. Soc. Brew. Chem., 60 (3), 130.<br />
SEWARD, B. J. (1992) Ferment, 5 (4), 275.<br />
SOLE, S. M. (2000) Ferment, 13 (4), 25.<br />
SOUTH, J. B. (1992) MBAA Tech. Quart., 29, 20.<br />
STARS, A. C., SOUTH, J. B. <strong>and</strong> SMITH, N. A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, p. 103.<br />
STOWELL, K. C. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 507.<br />
TAYLOR, D. G. <strong>and</strong> BOXALL, J. (1999) Ferment, 12 (6), 18.<br />
TENG, J., STUBITS, M. <strong>and</strong> LIN, E. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 47.<br />
VAAG, P., RIIS, P., KNUDSEN, A.-D., PEDERSEN, S. <strong>and</strong> MEILING, E. (1993) Proc. 24th Congr. Eur. Brew.<br />
Conv., Oslo, p. 155.<br />
WAINWRIGHT, T. (2003) Brewers' Guard., 132 (2), 20.<br />
WALKER, M. D., BOURNE, D. T. <strong>and</strong> WENN, R. V. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht,<br />
p. 191.<br />
WALKER, M. D. <strong>and</strong> WESTWOOD, K. T. (1991) J. Amer. Soc. Brew. Chem., 50, 4.<br />
WEBSTER, R. D. J. (1981) J. Inst. <strong>Brewing</strong>, 87, 52.<br />
WIEG, A. J. (1973) MBAA Tech. Quart., 10 (2), 7.<br />
WIEG, A. J. (1987) in <strong>Brewing</strong> Science, Vol. 3. (Pollock, J. R. A. ed.). p. 533. Academic Press, London.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
3<br />
Water, effluents <strong>and</strong> wastes<br />
3.1 Introduction<br />
Breweries use large amounts of water, (`liquor' in the UK). The actual amounts of water<br />
used ranging from three to (exceptionally) 30 times the volumes of beer produced. As<br />
beers usually have water contents of 91 98% (or even 89% in the cases of barley wines),<br />
<strong>and</strong> the amounts lost by evaporation <strong>and</strong> with by-products are relatively small it follows<br />
that large volumes of waste water are produced. Sometimes large volumes are produced<br />
because of operational inefficiencies but breweries operating in efficient but different<br />
ways, <strong>and</strong> with different product ranges, have substantially different water requirements.<br />
Apart from brewing, sparging <strong>and</strong> dilution liquors, water is used for a range of other<br />
purposes. These include cleaning the plant using manual or cleaning-in-place (CIP)<br />
systems, cooling, heating (either as hot water or after conversion into steam in a boiler),<br />
water to occupy the lines before <strong>and</strong> after running beer through them, for loading filter<br />
aids such as kieselguhr, for washing yeast <strong>and</strong> for slurrying <strong>and</strong> conveying away wastes<br />
as well as for washing beer containers such as tankers, kegs, casks <strong>and</strong> returnable bottles.<br />
The acquisition <strong>and</strong> treatment of liquor <strong>and</strong> the disposal of the brewery effluents are<br />
expensive processes <strong>and</strong> have long been studied.<br />
While water is the major component of beer the brewery takes in many other materials<br />
such as bottles <strong>and</strong> other packaging materials, malts, adjuncts <strong>and</strong> hops, <strong>and</strong> during the<br />
brewing <strong>and</strong> packaging processes `pollutants' <strong>and</strong> `wastes' are generated. These include<br />
broken glass, damaged cans, packaging materials such as cardboard <strong>and</strong> plastic, spent<br />
grains, spent hops, trub, tank bottoms, carbon dioxide, spilled or spoilt beer, wort, noise,<br />
odours, domestic wastes <strong>and</strong> heat. All these must be dealt with <strong>and</strong>, where possible,<br />
disposed of at a profit. This chapter is primarily concerned with the acquisition <strong>and</strong><br />
preparation of water of the grades needed in the brewery <strong>and</strong> the disposal of the dirty<br />
water, or effluents. However, the treatments or actions needed to deal with some other<br />
wastes or by-products are discussed (Anon., 1988; Armitt, 1981; Bak et al., 2001; Benson<br />
et al., 1997; Comrie, 1967; Crispin, 1996; Eden, 1987; Eumann, 1999; Grant, 1995;<br />
Hackstaff, 1978; Harrison et al., 1963; Hartemann, 1988; Heron, 1989; Mailer et al.,<br />
1989; Moll, 1979, 1995; Taylor, 1989; Theaker, 1988).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
3.2 Sources of water<br />
Thesourcesofavailablewatercanbeunderstoodwithreferencetothewatercycle.Water<br />
evaporates from the l<strong>and</strong>, plants, fresh water <strong>and</strong> the sea. In time this forms clouds <strong>and</strong><br />
precipitates as rain, snow or hail, falling back onto the l<strong>and</strong> or into the sea. Of that falling<br />
ontothel<strong>and</strong>aproportionevaporates,somerunsoffassurfacewater<strong>and</strong>somepenetrates<br />
intothesoil.Thesurfacewatermaybecollectedinlakes,riversorbehinddams<strong>and</strong>sobe<br />
available for use. Water from these sources is variously contaminated. Even rain-water is<br />
not pure, as it contains oxides of nitrogen <strong>and</strong> sulphur, dust, soot, pollen, microbes <strong>and</strong><br />
industrial wastes. Collected on the ground it may be further contaminated with industrial<br />
<strong>and</strong> domestic effluents, spillages, drainings from dumps, rotting plant materials, farm<br />
animal wastes, leached agricultural materials (fertilizers, pesticides, <strong>and</strong> herbicides) <strong>and</strong><br />
so on. The water which penetrates the soil is progressively filtered as it sinks downwards<br />
<strong>and</strong> so contains less of some surface-derived contaminants <strong>and</strong> micro-organisms. On the<br />
other h<strong>and</strong> salts may be dissolved from the pervious strata through which it passes. Thus<br />
surface waters will be comparatively `soft', i.e., will contain little in the way of dissolved<br />
salts, in contrast to waters recovered from underground, which may be either `soft' or<br />
`hard'. Water which passes through chalk or limestone becomes enriched with calcium<br />
bicarbonate, while in other areas it may contain calcium sulphate or salt. When the water<br />
meets an imperviouslayer the pervious layers above become saturated with water <strong>and</strong> are<br />
called aquifers. Water can be drawn from some of these. Near the sea the soil may be<br />
saturated with brine, with less dense fresh water layered above. In these areas fresh water<br />
must be withdrawn only slowly or the saline water table may be drawn up <strong>and</strong> brine will<br />
enter the well. Waters with very different characterstics may be available within one area<br />
(Rudin, 1976).<br />
Historically, different regions became famous for particular types of beer <strong>and</strong> in part<br />
these beer types were defined by the waters available for brewing (Table 3.1). Thus<br />
Pilsen, famous for very pale <strong>and</strong> delicate lagers has, like Melbourne, very soft water.<br />
Burton-on-Trent, with its extremely hard water, rich in calcium sulphate, is famous for its<br />
pale ales while Munich is well-known for its dark lagers, <strong>and</strong> Dublin (which has similar<br />
soft water) for its stouts. Breweries may receive water from different sources, which may<br />
be changed without warning. Water supplies may vary in their salt contents between day<br />
<strong>and</strong> night, from year to year <strong>and</strong> between seasons (Rudin, 1976; Byrne, 1990). It is now<br />
usual for breweries to adjust the composition of the water they use. In some few regions<br />
of the world saline water must be used, even sea water. In principle, several desalination<br />
methods might be used, but in <strong>practice</strong> it seems that purified water is obtained from sea<br />
water either by a highly thermally efficient distillation (Briggs et al., 1981), which is very<br />
costly, or by reverse osmosis (see below). Usually breweries obtain their water either<br />
from their own wells, springs or boreholes (surface waters are avoided where possible) or<br />
they may obtain them from water companies.<br />
Boreholes may extend downwards for 200 m (approx. 656 ft.), or more, <strong>and</strong> be fitted<br />
with an immersed pump to drive the water to the surface (Bak et al., 2001; Kunze, 1996).<br />
Water is drawn from an aquifer via a filter. A bore is sealed to prevent surface water or<br />
water from upper soil levels rapidly leaking down to the aquifer being used. While water<br />
from water companies is typically of a high st<strong>and</strong>ard of purity <strong>and</strong> is `potable', that is, it<br />
is fit for domestic use <strong>and</strong> is safe to drink, it is costly <strong>and</strong> is not necessarily fit to use in<br />
brewing (Baxter <strong>and</strong> Hughes, 2001). In addition, its composition <strong>and</strong> temperature are<br />
likely to vary <strong>and</strong> limits may be set on its use. Brewers' own water supplies will be more<br />
uniform, <strong>and</strong> will be substantially cheaper. However, there are likely to be charges for the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ight to abstract the water <strong>and</strong> the volumes <strong>and</strong> rates of abstraction will probably be<br />
limited to avoid exhausting the available ground water or seriously disturbing the water<br />
table.<br />
Most regions have strict regulations, which must be met before water is classified as<br />
being potable, <strong>and</strong> these provide the minimum st<strong>and</strong>ards for brewing waters (Armitt,<br />
1981; Bak et al., 2001; Baxter <strong>and</strong> Hughes, 2001; Moll, 1979, 1995). These regulations<br />
are often reviewed, the upper permitted limits for specified substances are frequently<br />
reduced <strong>and</strong> the numbers of substances mentioned are increased. Table 3.2 indicates how<br />
complex these `minimum st<strong>and</strong>ards' can be. The requirements may be grouped as<br />
`aesthetic' (colour, turbidity, odour <strong>and</strong> taste), microbiological st<strong>and</strong>ards (particularly the<br />
absence of pathogens), the levels of organic <strong>and</strong> inorganic materials that are in solution<br />
<strong>and</strong> the presence of radioactive materials. Some of these st<strong>and</strong>ards require comment.<br />
Drinking water must be safe, <strong>and</strong> so it must contain no pathogenic bacteria, protozoa, or<br />
viruses. However, the water is not necessarily sterile <strong>and</strong> so free of any organisms that<br />
can infect wort or beer, which must be the case for brewing water. The limits set for<br />
dissolved salts may be exceeded in some brewing waters. For example in Burton-on-<br />
Trent well waters the levels of calcium <strong>and</strong> sulphate ions may be very high (Table 3.1).<br />
Limitations on ammonia/ammonium <strong>and</strong> nitrogen levels are set since these are often<br />
indicators of contamination with decomposing organic matter. Nitrate levels, which vary<br />
widely, are a cause of concern as water sources are increasingly contaminated by nitrate<br />
from leached agricultural fertilizers. The fear is that during the preparation of the beer or<br />
in the consumer the nitrate may be reduced to nitrite (also limited, Table 3.2) <strong>and</strong> this, in<br />
turn, may give rise to carcinogens. The need to limit amounts of toxic ions is obvious<br />
although yeast needs trace amounts of many of them including copper, zinc, manganese<br />
<strong>and</strong> iron. These trace elements can be obtained from the brewers' grist. The minimum<br />
levels for total hardness <strong>and</strong> alkalinity are set to limit corrosion in pipework. Fluoride is<br />
often added to drinking water, but at the levels used it is harmless <strong>and</strong> without influence<br />
on fermentation.<br />
The organic contaminants mentioned (Table 3.2) deserve comment. Acrylamide, vinyl<br />
chloride <strong>and</strong> epichlorohydrin are toxic substances used in the manufacture of organic<br />
polymers <strong>and</strong> their presence indicates that unsafe disposal <strong>practice</strong>s have been used.<br />
Aldrin, dieldrin, heptachlor <strong>and</strong> heptachlor epoxide are insecticides or their metabolites.<br />
In other countries limits on other substances, including selective herbicides such as 2, 4-D<br />
(2, 4-dichloro phenoxyacetic acid) <strong>and</strong> diquat may be specified. Some of the polycyclic<br />
aromatic hydrocarbons are carcinogenic <strong>and</strong> the trihalomethanes confer unwanted<br />
flavours <strong>and</strong> may be toxic. The trihalomethanes, THMs, are unfortunately named since<br />
the organic substances in this group are not all based on the methane carbon skeleton <strong>and</strong><br />
not all are tri-substituted with halogens. Chlorine <strong>and</strong> bromine are the usual halogen<br />
substitutes (Cowan <strong>and</strong> Westhuysen, 1999; Grant, 1995; McGarrity, 1990; Taylor, 1989).<br />
THMs may be industrial solvent residues or they can arise from organic materials in the<br />
water when this is sterilized by chlorination. Thus they can be formed during water<br />
treatment in the brewery.<br />
Organic materials are particularly likely to be present in surface waters <strong>and</strong> may be<br />
dissolved or present as colloidal or suspended materials. Humic <strong>and</strong> fulvic acids are crude<br />
mixtures of organic materials with molecular weight ranges of 500 2,000,000 <strong>and</strong><br />
200 1,000 respectively. These are particularly likely to give rise to THMs during<br />
chlorination. The bromine substituents can be added when the chlorinated water contains<br />
bromide ions. The composition of the THM group varies. It includes chloroform,<br />
bromomethane, carbon tetrachloride, 1, 1-dichloroethane, 1, 1, 2-trichloroethane <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 3.1 Analyses of some waters from famous brewing centres, (expressed as mg/l). The analyses of these, or any waters do not remain constant with time<br />
(Moll, 1995; Mailer et al., 1989)<br />
Parameter Pilsen Burton-on-Trent MuÈnchen (Munich) Dortmund London Wien (Vienna) Melbourne<br />
Total dry solids 51 ± 1226 536 273 984 320 984 25<br />
Calcium (Ca 2+ ) 7.1 352 268 109 80 237 90 163 1.3<br />
Magnesium (Mg 2+ ) 3.4 24 62 21 19 26 4 68 0.8<br />
Bicarbonate (HCO3 ) 14 320 ± 171 ± 174 ± 243 ±<br />
Carbonate (CO3 2 ) ± ± 141 ± 164 ± 123 ± 3.6<br />
Sulphate (SO 4 2 ) 4.8 820 638 7.9 5 318 58 216 0.9<br />
Nitrate (NO3 ) tr. 18 31 53 3 46 3 tr. 0.2<br />
Chloride (Cl ) 5.0 16 36 36 1 53 18 39 6.5<br />
Sodium (Na + ) ± ± 30 ± 1 ± 24 ± 4.5<br />
tr. ˆ Traces.<br />
± ˆ Not given.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 3.2 A list of the maximum (minimum) concentrations of substances that may not be<br />
exceeded in drinking water in the UK in 2001 (courtesy of J. MacDonald). Compare Bak et al.,<br />
(2001); Baxter <strong>and</strong> Hughes (2001)<br />
Parameter Units Concentration or value<br />
Colour mg/l (Pt/Co scale) 20<br />
Turbidity Formazin units 1<br />
Odour Dilution number 3 at 25 ëC<br />
Taste Dilution number 3 at 25 ëC<br />
Temperature ëC 25<br />
pH (limits) pH units 6.5±10.0<br />
Conductivity S/cm at 20 ëC 2500<br />
Permanganate value O2, mg/l 5<br />
Total organic carbon, TDC C, mg/l no significant increase<br />
Total coliform bacteria number/100 ml 0<br />
Faecal coliform bacteria number/100 ml 0<br />
Faecal Streptococci, Enterococci number/100 ml 0<br />
Clostridium perfringens number/100 ml 0<br />
Sulphate reducing Clostridia number/20 ml 1<br />
Colony counts number/ml at 25 or 37 ëC no significant increase<br />
(In some regions tests are also<br />
carried out for Protozoa, such as<br />
Cryptosporidium <strong>and</strong> Guiardia)<br />
Radioactivity (total indicative dose) MSv/year 0.1<br />
Tritium Bq/l 100<br />
Boron B mg/l 1<br />
Chloride Cl, mg/l 250<br />
Calcium Ca, mg/l 250<br />
Total hardness Ca, mg/l 60 (minimum)<br />
Alkalinity HCO 3, mg/l 30 (minimum)<br />
Sulphate SO4, mg/l 250<br />
Magnesium Mg, mg/l 50<br />
Sodium Na, mg/l 200<br />
Potassium K, mg/l 12<br />
Dry residues (after 180 ëC) mg/l 1500<br />
Nitrate NO3, mg/l 50<br />
Nitrite NO2, mg/l 0.5<br />
Ammonia, ammonium ions NH 4, mg/l 0.5<br />
Kjeldahl nitrogen N, mg/l 1.0<br />
Dissolved or emulsified hydrocarbons<br />
Mineral oils g/l 10<br />
Benzene g/l 1<br />
Phenols C6H5OH, g/l 0.5<br />
Surfactants (detergents) as lauryl sulphate, g/l 200<br />
Aluminium Al, g/l 200<br />
Iron Fe, g/l 200<br />
Manganese Mn, g/l 50<br />
Copper Cu, mg/l 2<br />
Zinc Zn, mg/l 5<br />
Phosphate P, mg/l 2.2<br />
Fluoride F, mg/l 1.5<br />
Silver Ag, g/l 10<br />
Arsenic As, g/l 10<br />
Bromate BrO3, g/l 10<br />
Cadmium Cd, g/l 5<br />
Cyanide CN, g/l 50<br />
Chromium Cr, g/l 50<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 3.2 Continued<br />
Parameter Units Concentration or value.<br />
Mercury Hg, g/l 1<br />
Nickel Ni, g/l 20<br />
Lead Pb, g/l (will be reduced<br />
in 2013) 25<br />
Antimony Sb, g/l 5<br />
(Elsewhere limits are set on other substances, such as thallium, beryllium, uranium <strong>and</strong> asbestos)<br />
Acrylamide g/l 0.1<br />
Vinyl chloride g/l 0.5<br />
Epichlorohydrin g/l 0.1<br />
Aldrin g/l 0.03<br />
Dieldrin g/l 0.03<br />
Heptachlor g/l 0.03<br />
Heptochlorepoxide g/l 0.03<br />
Other pesticides g/l 0.1<br />
Pesticides, total g/l 0.5<br />
Polycyclic aromatic hydrocarbons* g/l 0.1<br />
Benzo(a)-3,4-pyrene ng/l 10<br />
1,2-Dichloroethane g/l 3<br />
Tetrachloromethane g/l 3<br />
Trichloroethane g/l 10<br />
Tetrachloroethane & trichloroethene g/l 10<br />
Trihalomethanes, total y g/l 100<br />
Substances extractable in chloroform mg/1, dry residue 1<br />
* Sum of individual concentrations of members of a list of substances benzo[b]fluoranthene, benzo[k]fluoranthene,<br />
benzo-11,12-fluoranthene, benzo[ghi]perylene <strong>and</strong> indeno-[1,2,3-cd]pyrene.<br />
y Sum of chloroform, bromoform, dibromochloromethane <strong>and</strong> dibromodichloromethane.<br />
tetrachloroethane together with arange of other substances. Some THMs are also VOCs,<br />
(volatile organic compounds). Their volatility is the basis of their partial or total removal<br />
during gas-stripping processes (as when removing carbon dioxide after dealkylation, or<br />
oxygen removal) or during mashing <strong>and</strong> in the copper boil. THMs are also removed by<br />
active carbon filtration <strong>and</strong> partly removed during reverse osmosis. Chlorination of<br />
aromatic, organic materials can give rise to other undesirable materials, including<br />
medicinally flavoured chlorophenols.<br />
3.3 Preliminary water treatments<br />
Most brewers find it necessary to treat the water coming into the brewery. The variety of<br />
substances that may occur in water is large, <strong>and</strong> different treatments are needed to deal<br />
with them (Fig. 3.1). Different waters require different treatments <strong>and</strong> brewers require<br />
grades of water treated in different ways depending on the uses to which it will be put. In<br />
some instances it may only be necessary to pre-treat the liquor, while in other cases<br />
extensive further treatment will be needed. Preliminary treatments may involve aeration,<br />
sedimentation (with or without the prior addition of coagulants <strong>and</strong> flocculating agents,<br />
which initially require vigorous stirring, followed by more gentle stirring to encourage<br />
the build up of flocs), flotation, filtration <strong>and</strong> sterilization. Some of these treatments may<br />
be used more than once (e.g. sterilization) during the preparation of liquors. Aeration<br />
with compressed air (with or without ozone) is used to oxidize ferrous ions to ferric<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Diameter (μm)<br />
Types of substance<br />
Water treatments<br />
10 –4<br />
10 –3<br />
10 –2<br />
10 –1<br />
1 10 10 2<br />
10 3<br />
Ions <strong>and</strong> molecules<br />
in true solution<br />
Macromolecules Microparticles Fine particles<br />
Salts<br />
Sugars<br />
Colloids<br />
Suspended solids<br />
Viruses Bacteria<br />
Pollen<br />
S<strong>and</strong><br />
Yeasts<br />
Biological uptake<br />
Chemical precipitation<br />
Particle filtration Screening<br />
Ion exchange<br />
Microfiltration<br />
Sedimentation<br />
Ultrafiltration<br />
S<strong>and</strong> filtration <strong>and</strong><br />
Nanofiltration<br />
microscreening<br />
Reverse osmosis<br />
Adsorption<br />
Coagulation <strong>and</strong> filtration<br />
Fig. 3.1 The sizes of dissolved, colloidal <strong>and</strong> suspended materials that must be considered in<br />
water purification <strong>and</strong> the methods used in removing them (Bak et al., 2001; Briggs et al., 1981).<br />
oxide/hydroxide (which separates from solution), to remove volatile organic substances,<br />
hydrogen sulphide, <strong>and</strong> carbon dioxide from water. Water is sprayed onto the top of a<br />
column filled with plastic packing, <strong>and</strong> flows downwards against a counter-current flow<br />
of air, which carries away the unwanted volatile substances.<br />
Measured amounts of ferric chloride or aluminium sulphate, with or without some<br />
organic polyelectrolyte, may be added to water to act as coagulants. The salts hydrolyse,<br />
giving rise to voluminous precipitates of hydrated ferric hydroxide or aluminium<br />
hydroxide. After thorough mixing the precipitates are allowed to settle, carrying down<br />
inorganic <strong>and</strong> some organic suspended matter. The comparatively clear supernatant is<br />
removed, leaving the sludge to be collected <strong>and</strong> dumped. Untreated water may also have<br />
a sedimentation treatment to allow the denser suspended materials to settle, or it may be<br />
filtered. If the water is rich in dissolved iron or manganese these should be removed. Iron<br />
in particular can deposit oxides as slime which blocks pipes <strong>and</strong> can clog filters. In the<br />
brewing process iron ions give colours with polyphenols <strong>and</strong>, probably acting as<br />
oxidation catalysts, promote flavour <strong>and</strong> haze instability. Aeration or treatment with<br />
oxidizing agents, such as chlorine, converts ferrous ions to ferric ions which then separate<br />
as ferric hydroxide. Oxidizing agents are also needed to convert manganese to a<br />
precipitable form. Sedimentation or flotation are generally used before filtration. To<br />
achieve sedimentation the water is passed into a large tank in which it moves quietly <strong>and</strong><br />
slowly to allow the solids to precipitate or the water may pass through lamellar<br />
separators, or centrifuges or hydrocyclones (Bak et al., 2001). Alternatively, flocculated<br />
material may be removed by flotation in which finely divided bubbles of air rise from the<br />
base of a vessel <strong>and</strong> carry the flocs to the surface, where they accumulate <strong>and</strong> are<br />
removed by skimming.<br />
Often suspended materials are removed from water by coarse filtration. It may be<br />
passed through a bed of sharp, calcined s<strong>and</strong> that may be 2 3 m (6.6 10 ft.) thick or it<br />
may pass through successive layers of granular plastic (3 5 mm), anthracite (2 3 mm)<br />
<strong>and</strong> s<strong>and</strong> (0.5 1.5 mm). When the filter becomes blocked, as signalled by a rising<br />
resistance to the water flow, it is back-flushed with a reverse stream of water to carry<br />
away the blocking particles. In special BIRM filters the s<strong>and</strong> is mixed with manganese<br />
dioxide, which catalyses the oxidation, <strong>and</strong> so the precipitation, of ferrous ions as ferric<br />
hydroxide. A newer device is the fibrous depth filter. Fibres are firmly twisted together<br />
around a support to form a tube, creating an efficient filter, which is able to exclude more<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
than 98% of particles over 2 m in diameter. To clean the filter the tension is reduced <strong>and</strong><br />
the enlarged spaces between the fibres are cleaned with a back-wash. Cartridge filters,<br />
which exclude particles over 5 m, may also be used.<br />
Water in breweries may be sterilized more than once at different stages. Chemical<br />
sterilants are chlorine, hypochlorites, chlorine dioxide, ozone <strong>and</strong>, less often, silver.<br />
Physical sterilants used are exposure to ultraviolet light, sterilizing filtration <strong>and</strong>, rarely<br />
(except during the hop-boil), heat. Chlorine, used as the green-yellow gas or as sodium,<br />
potassium, or calcium hypochlorites, is a commonly used sterilant. One recommendation<br />
is that the level of available chlorine should initially be 5 mg/l that the water should be<br />
held at least for 30 min., to allow the sterilant to act <strong>and</strong> at this time the level of free<br />
chlorine should not have fallen below 1 mg/l. This recommendation emphasizes that<br />
chlorine is a highly reactive compound <strong>and</strong> a strong oxidizing agent that is used up during<br />
water sterilization but that it remains in solution long enough to have a useful `residual'<br />
sterilizing effect. Disadvantages of using chlorine include the formation of chlorophenols<br />
<strong>and</strong> THMs from organic substances in the water. Unwanted residual chlorine may be<br />
removed by aeration, evaporation, by filtration through active carbon, or by adding<br />
bisulphite or sulphite to the liquor, when the chlorine is reduced to chloride ions while the<br />
sulphite is oxidized to sulphate. If the water contains ferrous ions chlorine will oxidize<br />
them to ferric ions, which will then form flocculent ferric hydroxide.<br />
Chlorine dioxide, ClO2, an unstable, yellow, explosive gas that is generated on site<br />
immediately before use, from hydrochloric acid <strong>and</strong> sodium chlorite:<br />
5NaClO4 ‡ 4HCl ! 4ClO2 ‡ 5NaCl ‡ 2H2O<br />
It is a strong oxidizing agent. Unlike chlorine, it does not chlorinate organic substances<br />
<strong>and</strong> so does not give rise to THMs or unwanted flavour compounds. Indeed it destroys the<br />
off-flavours given by some chlorophenols. Its `residuals' last for a shorter period than<br />
those of chlorine <strong>and</strong> so this agent is less effective at preventing re-infection. A contact<br />
time of 15 min. is desirable. Ozone, O 3, is formed on site by passing dry air or oxygen<br />
through an electrical generator. This gas is a strong oxidizing agent, but its lifetime is<br />
short <strong>and</strong> so it gives almost no residual protection against re-infection. This agent is said<br />
to be more effective against Giardia, cysts of other protozoa <strong>and</strong> some viruses <strong>and</strong><br />
bacteria than chlorine or chlorine dioxide. It is also effective at destroying some taints<br />
<strong>and</strong> odours. Treated water should initially contain 1 3 g ozone/m 3 . Treatment should be<br />
extended from 3 to 15 min. <strong>and</strong> the higher doses should be used if iron <strong>and</strong>/or manganese<br />
ions are to be oxidized. Ozone is toxic <strong>and</strong> should be degraded before waste gases are<br />
vented to the atmosphere. Silver ions, generated by the electrolytic `katadyn process' are<br />
also effective sterilants under some conditions but their use is not permitted everywhere<br />
<strong>and</strong> impurities in the water can reduce their effectiveness. All the sterilants mentioned<br />
must be h<strong>and</strong>led with care, as they can be dangerous.<br />
Sterilization with ultraviolet light, UV, relies on the fact that emissions at wavelengths<br />
around 260 nm are absorbed by nucleic acids, which are then disrupted. Thus UV light from<br />
low-pressure mercury lamps is able to kill microbes, including viruses, but of course the<br />
treatment leaves no sterilizing residue. The long tubular lamps are housed in quartz tubes<br />
<strong>and</strong> the water flows past in a tubular metal housing which limits the length of the UV light<br />
path. The water must be clear <strong>and</strong> colourless <strong>and</strong> not give deposits to avoid blocking the<br />
UV radiation. The dwell time in the radiation chamber must be sufficient for sterilization to<br />
be complete. UV treatment of water containing dissolved ozone is more effective than<br />
either agent alone, <strong>and</strong> chlorinated hydrocarbons are fully oxidized, to carbon dioxide <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
hydrochloric acid, <strong>and</strong> the COD of the water is reduced. The lamp tubes must be checked<br />
regularly <strong>and</strong> must be replaced as they approach the end of their working lives. Personnel<br />
must not be exposed to this radiation. Bacteria <strong>and</strong> fungi (but not viruses) can be removed<br />
by sterile filtration through special membranes (wound membranes or hollow fibres)<br />
having, for example, notional pore sizes of 0.2 or 0.45 m. Such membranes can easily be<br />
blocked <strong>and</strong> so the water must be free of components that can deposit sludge or scale or<br />
contain fine suspended matter therefore the water to be sterilized must be pre-treated <strong>and</strong><br />
carefully filtered. Pasteurization is rarely applied to water, but is used on some beers<br />
(Chapter 21). Other treatments, such as flocculation or reverse osmosis, deplete or remove<br />
microbes, but these processes are primarily used for other purposes.<br />
3.4 Secondary water treatments<br />
Waterusedinbreweriesisusuallytreatedtoadjustitscomposition(Baketal.,2001;Benson<br />
etal.,1997;Blackmann,1998;Comrie,1967;Maileretal.,1989;Moll,1995;Taylor,1989).<br />
Treatments may reduce levels of organic compounds in solution or adjust the ionic<br />
compositionoftheliquor.Inthepastthissubjectwasconfusedbywidelydifferingmethods<br />
of expressing salt concentrations (Moll, 1979,1995; Appendix). Here units of mg ion/l will<br />
be used. Ions in beer can influence its flavour (see below) <strong>and</strong> calcium ions in particular<br />
influence the mashing process (Chapter 4). Discussions of water composition often involve<br />
the term `hardness'. `Soft' water contains low concentrations of dissolved salts, particularly<br />
salts of calcium <strong>and</strong> (with less emphasis) salts of magnesium. `Hard' water contains high<br />
concentrations of salts, usually mainly calcium bicarbonate or calcium sulphate. `Temporary<br />
hardness' is caused chiefly by calcium bicarbonate <strong>and</strong> is so-called because if the water is<br />
boiled the bicarbonate is converted to the carbonate, which precipitates leaving the clarified<br />
water `softened'. In contrast `permanent hardness' is mainly caused by calcium sulphate,<br />
<strong>and</strong> this remains in solution when the water is boiled. The distinction is important if the<br />
liquor is to be used for mashing or, even more, for sparging.<br />
While temporary hardness can be removed by boiling water, this process is costly <strong>and</strong><br />
is usually avoided although it may be beneficial in other ways, such as sterilizing the<br />
water, driving out the dissolved oxygen <strong>and</strong> evaporating volatile contaminants such as<br />
THMs. The decomposition of the bicarbonate occurs as:<br />
Ca(HCO 3† 2 ! CaCO3 # ‡ H2O+CO2 "<br />
Magnesium bicarbonate is also decomposed by boiling, but magnesium carbonate is<br />
appreciably soluble, <strong>and</strong> so its removal is incomplete. The calcium carbonate precipitates<br />
<strong>and</strong> the carbon dioxide is driven off. Boiling also accelerates the oxidation of ferrous ions<br />
to ferric ions, which precipitate as the hydroxide. Treatments with lime water may be<br />
used to remove temporary hardness from water. A calculated amount of lime-water, or a<br />
slurry of lime in water, is mixed with the water. Calcium carbonate is precipitated:<br />
Ca…HCO3† 2 ‡ Ca…OH† 2 ! 2CaCO3 # ‡ 2H2O<br />
CO2 ‡ Ca…OH† 2 ! CaCO3 # ‡ H2O<br />
In older plant the precipitate of calcium carbonate settles slowly, but in a more modern<br />
<strong>and</strong> fully automated plant the calcium carbonate is deposited on crystalline granules of<br />
the same material 0.1 2.5 mm in diameter. In either case residual suspended calcium<br />
carbonate is removed, for example by s<strong>and</strong> filtration. The calcium carbonate is used in<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
agriculture, spread on fields to reduce soil acidity. After the lime treatment the water is<br />
alkaline <strong>and</strong>, for brewing purposes, must be adjusted to about pH 7. In Germany, <strong>and</strong><br />
elsewhere where the use of mineral acids is forbidden, this is achieved by adding carbon<br />
dioxide. However, where it is permitted, the alkalinity is reduced by additions of food<br />
grade acids, commonly mineral acids but sometimes lactic acid. When the water contains<br />
an appreciable amount of magnesium bicarbonate it may receive a `split treatment'. A<br />
portion of the water is dosed with a high level of lime. Calcium carbonate precipitates <strong>and</strong><br />
magnesium precipitates as the hydroxide:<br />
Mg…HCO3† 2 ‡ 2Ca…OH† 2 ! Mg…OH† 2 # ‡2CaCO3 # ‡H2O<br />
This partly treated water is strongly alkaline, with a pH of about 12. It is mixed with the<br />
remainder of the water (about two-thirds of the amount being treated), precipitating the<br />
calcium bicarbonate as the carbonate. After clarification the pH of the water is adjusted.<br />
Thus calcium bicarbonate <strong>and</strong> some of the magnesium salts are removed. Both the lime<br />
water treatments also precipitate iron <strong>and</strong> manganese ions, as hydroxides, <strong>and</strong> the<br />
precipitates entrain <strong>and</strong> remove some organic contaminants. Another way of removing<br />
bicarbonate ions from solution is to acidify the water <strong>and</strong> then remove the carbon dioxide<br />
formed with aeration. Thus:<br />
Ca…HCO3† 2 ‡ H2SO4 ! CaSO4 ‡ 2 H2CO3<br />
H2CO3 ! H2O ‡ CO2 "<br />
The food-grade acid used depends on flavour, safety <strong>and</strong> operational considerations.<br />
After acidification the water is passed down a packed tower against an upward stream of<br />
air that carries away the carbon dioxide. Incidentally, it also removes some volatile<br />
organic compounds <strong>and</strong> chlorine, if these are present.<br />
Several types of ion exchange treatments may be applied to brewing waters. Modern<br />
ion exchange resins are now used rather than the old mineral ion exchangers, such as<br />
zeolites. The resins are beads of varying porosities, often of cross-linked polystyrene,<br />
which carry acidic or basic groups. Ion exchange treatments may be fully automated.<br />
Resins must be free of flavoured, low molecular weight organic materials, <strong>and</strong> they must<br />
not be exposed to chlorine, which will attack them. Iron <strong>and</strong> manganese must have been<br />
removed from the feed water <strong>and</strong> this is carefully filtered to prevent the resin beds<br />
becoming blocked. The costs of ion exchange treatments include the costs of regenerating<br />
the resins <strong>and</strong> of disposing of the regeneration liquid chemical wastes. The treatments<br />
may be divided into dealkalization, softening <strong>and</strong> demineralization. In dealkalization,<br />
which removes temporary hardness, the water is passed through a packed column of a<br />
weakly acidic cation exchange resin, which carries carboxylic acid groups. This resin<br />
exchanges hydrogen ions for calcium <strong>and</strong> magnesium ions in the water. The hydrogen<br />
ions combine with bicarbonate ions in the water forming carbonic acid <strong>and</strong> this then<br />
dissociates reversibly to carbon dioxide <strong>and</strong> water:<br />
Res 2 2H 2 ‡ Ca 2‡ ! Res 2 Ca 2‡ ‡ 2H ‡<br />
H ‡ ‡ HCO3 $ H2CO3 $ CO2 " ‡ H2O<br />
The water is then passed down an aeration tower where the carbon dioxide is removed<br />
together with some volatile organic compounds, VOCs, <strong>and</strong> chlorine.<br />
After ion exchange treatment water is often passed through active carbon filters as a<br />
precaution to remove any unwanted off-flavoured compounds that may be released from<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the resins. In time the exchange capacity of the resin is exhausted <strong>and</strong> it has to be<br />
regenerated. Both the resin <strong>and</strong> the acid used to regenerate it are comparatively<br />
inexpensive <strong>and</strong> indeed this treatment may be used before demineralization to reduce the<br />
cost of this latter process. The waste regeneration liquid is acidic. Water softening can be<br />
carried out by adding sodium carbonate to water containing, for example, calcium<br />
sulphate. Calcium carbonate precipitates leaving the more soluble sodium sulphate in<br />
solution. More often softening is carried out by ion exchange. A strongly acidic ion<br />
exchange resin, carrying sulphonic acid groups, is loaded with sodium ions. When the<br />
water passes through the resin this exchanges sodium ions for the more strongly bound<br />
divalent metal ions, like those of calcium <strong>and</strong> magnesium. The softened water is used<br />
where the use of hard water might give rise to scales <strong>and</strong> deposits of sludge, for example<br />
in cooling water, in boilers <strong>and</strong> in rinsing water.<br />
Demineralization involves treating the water with strongly acidic <strong>and</strong> strongly basic<br />
resins loaded with hydrogen <strong>and</strong> hydroxyl ions respectively. The water may go through<br />
two resin beds working in series or mixed bed resins may be used. Aeration may be used,<br />
either after passage through the strongly acidic resin or after the entire treatment, to<br />
remove liberated carbon dioxide. Mixed bed resins are returned to the makers for<br />
regeneration. Using this system all the positively charged ions in the water are exchanged<br />
for hydrogen ions <strong>and</strong> all the negatively charged ions are exchanged for hydroxyl ions.<br />
For example:<br />
Res 2‡ 2OH ‡ SO4 2 ! Res 2‡ SO4 2 ‡ 2OH<br />
Res 2 2H ‡ ‡ Ca 2‡ ! Res 2 Ca 2‡ ‡ 2H ‡<br />
H ‡ ‡ OH $ H2O<br />
The hydrogen <strong>and</strong> hydroxyl ions combine to give water. Demineralization can give very<br />
pure water. Very strong basic resins can even remove silicate ions <strong>and</strong> some organic acids<br />
<strong>and</strong> the resins can at least partly remove some herbicides <strong>and</strong> their breakdown products.<br />
Temporary <strong>and</strong> permanent hardnesses are removed <strong>and</strong> so are all ions, including nitrate.<br />
It is now commonplace for brewing water to be demineralized <strong>and</strong> then for the<br />
compositions of the process streams to be adjusted to meet the different process<br />
requirements. This is convenient, since fluctuations in the composition of the incoming<br />
water become irrelevant <strong>and</strong> different brewing liquors can be prepared as needed for<br />
different beers. The processes of demineralization <strong>and</strong> reverse osmosis are in direct<br />
competition. If the levels of the total dissolved salts are comparatively low (TDS < 1,000<br />
mg/l) then demineralization is likely to be chosen, despite the cost of regenerating the<br />
resins. Typically some 10 15% of the incoming water is used in the regeneration<br />
processes <strong>and</strong> goes to waste.<br />
Reverse osmosis (RO) is attractive if the water to be purified is high in total dissolved<br />
solids (> 1,000 mg/l; Benson et al., 1997; Berkmortel, 1988a. 1988b; McGarrity, 1990;<br />
Thompson, 1995). There have been great advances in the technology of making<br />
membrane units either in the form of hollow fibres or as spirally wound sheets. The semipermeable<br />
membranes used in reverse osmosis are permeable to water but they are<br />
impermeable to microbes, ions <strong>and</strong> organic substances of molecular weight > 200. If pure<br />
water is separated from a salt solution by a semi-permeable membrane (i.e. permeable to<br />
water but not to solutes) there is a net migration of water through the membrane into the<br />
salt solution. If the pressure on the salt solution is increased then at a particular value it<br />
can balance the osmotic pressure, the tendency of the water to migrate into the salt<br />
solution, <strong>and</strong> no net movement of water will occur. If the pressure on the salt solution is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
increased still more water will be driven through the membrane from the salt solution,<br />
which is concentrated, <strong>and</strong> the permeate will be substantially pure water. High-pressure<br />
pumpsareneededtodrivethisprocess.Inanextremecaseapressureof25bar(367.5psi)<br />
is needed to desalinate sea water with aTDS of 35,000mg/l. This extreme degree of<br />
desalination may be carried out in two steps. The water under pressure flows across the<br />
membranes <strong>and</strong> about 75% is recovered as purified permeate <strong>and</strong> 25% as concentrated<br />
`saline'. In many cases this `saline water' is still of use for hosing down, etc.<br />
InalessextremecaseafeedwaterwithaTDSof1200mg/ltreatedbyreverseosmosis<br />
(RO) gave risetoapermeate with 1 8mg/l TDS.RO plant hasno regenerationcosts<strong>and</strong><br />
can be cleaned automatically (CIP). On the other h<strong>and</strong> many of the membranes must be<br />
protected from chlorine <strong>and</strong>, to prevent membrane blockages, the feed water must be free<br />
of suspended solids, manganese or iron salts or materials that can form scales or sludges.<br />
Therefore the water may need pre-treatment <strong>and</strong> it must be filtered, probably through a<br />
s<strong>and</strong> filter <strong>and</strong> then through afine filter removing particles >10 mdiameter (Braun <strong>and</strong><br />
Niefind, 1988). Filtration also protects the high-pressure pumps from damage byabrasive<br />
particles. As the salt concentration of the water increases so does the cost of treatment,<br />
but not to aproportional extent. Sets of membranes are expected to last about five years<br />
therefore atreatment could involve filtration through aBIRM filter, acidification, the<br />
removal of carbon dioxide in an aeration tower, very fine filtration <strong>and</strong> then RO. Reverse<br />
osmosis is now in widespread use. In contrast, electrodialysis, acompeting technology,<br />
seems not to have found favour.<br />
<strong>Brewing</strong> water is often passed through layers of active carbon. This `carbon filtration'<br />
is used to remove residual chlorine, humic <strong>and</strong> fulvic acids, many aromatic organic<br />
substances, some pesticides, some THMs <strong>and</strong> phenolic substances, <strong>and</strong> unwanted<br />
coloured, flavoured <strong>and</strong> odorous materials. Charcoals from different sources differ in<br />
their adsorbtive capacities <strong>and</strong> the types of substances that they remove best (Gough,<br />
1995). Bituminous coal, anthracite or coconut shells, as examples, are pyrolysed, giving<br />
products that are predominantly microcrystalline graphite. These are then `activated' by<br />
one of several methods. The material chosen for use must have the correct particle sizes,<br />
be strong enough to resist some wear, be of a `food grade', <strong>and</strong> have the correct<br />
adsorbtive characteristics. Acharge for afilter should last for five to seven years. The<br />
liquorreachingthefiltershouldbesterile<strong>and</strong>wellfiltered<strong>and</strong>havebeentreatedsothatit<br />
does not give deposits. With the passage of time the filter will tend to become blocked<br />
<strong>and</strong> asource of microbiological infection. In addition, its adsorbtive capacity will tend to<br />
become saturated <strong>and</strong> so the charcoal must be cleaned, sterilized <strong>and</strong> regenerated. As a<br />
routine, carbon filters are backwashed with chlorinated water <strong>and</strong> then drained <strong>and</strong><br />
steamed. Sterilization of the liquor coming from acarbon filter must be by atechnique<br />
that leaves no residues. Consequently UV radiation is often used or, less often, ozone.<br />
Brewersareincreasinglyconcernedtoexcludeair,orrathertheoxygenintheair,from<br />
their beers <strong>and</strong> from the production stream. To help to achieve this several methods for<br />
deoxygenating water are in use (Andersson <strong>and</strong> Norman, 1997; Benson et al., 1997;<br />
Cleather, 1992; Kunze, 1996). In some instances the carbon dioxide <strong>and</strong>/or nitrogen<br />
levels of the liquor are adjusted as the oxygen is removed. As the temperature of water<br />
rises so the amount of oxygen that it will hold in solution declines (Table A12 on page<br />
844) therefore water can be at least partly deaerated by boiling or stripping with (clean)<br />
steam. This approach is costly. Another method is to heat the water to 85 ëC (185 ëF) or<br />
more <strong>and</strong> then spray it into a vacuum chamber to remove the dissolved gas. More than<br />
one treatment may be needed. Another approach is to saturate, under pressure, the water<br />
to be deoxygenated with carbon dioxide or nitrogen. The water is then transferred into a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
low-pressure chamber when the gas bursts out of solution carrying the dissolved oxygen<br />
with it. The chamber may be ventilated with the carrier gas. Merely bubbling carbon<br />
dioxide or nitrogen through the water is not sufficient.<br />
Several methods seem to be preferable to those described; water can be sprayed into<br />
the top of atower packed with aplastic filler. As the water trickles down as athin film it<br />
meets an upflow of the stripping gas, carbon dioxide or nitrogen. The stripping gas may<br />
be sterile filtered. The liquor loses its oxygen to the stripping gas, which removes it<br />
efficiently as the system works on the counter-current principle. An alternative approach<br />
is to pass the water across ahydrophobic but gas-permeable membrane (hollow fibre or<br />
spirally wound format. Brown et al., 1999). The non-water side of the membrane may be<br />
avacuum or aflowing stripping gas. The water loses its oxygen through the membrane.<br />
Several of the methods mentioned ensure that the water is saturated with the chosen<br />
carrier gas. Another method for reducing dissolved oxygen, DO, levels is its catalytic<br />
reduction with hydrogen. The DO level of astream of flowing water is sensed <strong>and</strong> the<br />
appropriate amount of hydrogen is dosed into the water stream <strong>and</strong> is dissolved under<br />
pressure. After thorough mixing the water passes over apalladium catalyst (supported on<br />
beads of an ion exchange resin) <strong>and</strong> the oxygen is reduced to water. Using this technique<br />
the oxygen content of the water can be reduced to as little as 0.002mgO2/l. In some<br />
instances, e.g., in some boiler waters, sulphite salts may be added to react with, <strong>and</strong> so<br />
`scavenge', oxygen. The sulphite is oxidized by the oxygen, becoming sulphate.<br />
3.5 Grades of water used in breweries<br />
The mixture of dissolved substances judged suitable in liquor used for mashing may<br />
differ from those present in sparge liquor or dilution liquor <strong>and</strong> will certainly differ from<br />
those preferred in cleaning or boiler waters. Mashing liquor may not be completely<br />
sterile, but its microbial count must be low. Increasingly, brewers employing newer types<br />
of plant will mash with oxygen-reduced or oxygen-free water <strong>and</strong> under conditions such<br />
that oxygen pick-up is minimal. The mixture of salts present in the liquor may have been<br />
supplemented or adjusted. The addition of calcium sulphate <strong>and</strong>/or chloride, `Burtonization',<br />
is common <strong>and</strong> when demineralized or reverse osmosis processed water is used as<br />
the base allthe saltspresent willhave been added. Thesaltsused <strong>and</strong>their concentrations<br />
are decided with reference to their functions in mashing (Ch. 4; cf. Table 3.1) <strong>and</strong> their<br />
flavours. To reduce the pH of a malt mash by 0.1 unit requires the addition of 300 g<br />
calcium sulphate or 250 g calcium chloride/100 kg malt, the salts being added in the<br />
brewing <strong>and</strong> sparging liquor (Comrie, 1967).<br />
The flavours of chloride <strong>and</strong> sulphate ions are different. It is recommended that brewing<br />
liquors for Burton style pale ales should have a sulphate to chloride ratio of 2:1 to 3:1. For<br />
mild ales the concentration of calcium should be less <strong>and</strong> the ratio should be about 2:3,<br />
while liquor for stouts should contain little or no sulphate. Sparge liquors may resemble<br />
mashing liquors, but it is desirable that the bicarbonate levels are very low, otherwise there<br />
is an undesirable rise in the pH of the last runnings as the buffering substances are leached<br />
from the mash. Deaerated <strong>and</strong> sterile water is required for pre-run <strong>and</strong> chase water<br />
preceding <strong>and</strong> following beer through pipework <strong>and</strong> for carrying slurried kieselguhr when<br />
forming the pre-coat on filters. Water that is sterile, deoxygenated, correctly carbonated <strong>and</strong><br />
has the correct ionic composition <strong>and</strong> pH is used to dilute `high gravity' beers to their final<br />
strengths. Sterile water is also used to slurry <strong>and</strong> wash yeast. When water is to be heated<br />
during use, as in cooling water or in the pasteurizer, it needs to have been softened,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
demineralized or otherwise treated to prevent the deposition of sludges <strong>and</strong> scale, which<br />
can cover surfaces <strong>and</strong> interfere with heat exchange <strong>and</strong> may even block pipework. In<br />
addition, carbon dioxide <strong>and</strong> oxygen should be removed to minimize corrosion <strong>and</strong><br />
antimicrobial agents may be added. In some instances the pH of heating water is adjusted<br />
with phosphate salts <strong>and</strong> scale-softening agents, such as tannins, may be added.<br />
While some high-pressure boilers require asupply of fully de-ionized, oxygen-free<br />
water, low-pressure boilers may operate with softened water sometimes dosed with<br />
chelating agents, such as EDTA or polyphosphates, to prevent the deposition of calcium<br />
salts on the heat-exchange surfaces (Ibbotson, 1986). Sludges are removed by `blowingdown',<br />
that is, ejecting them from the boiler to waste. Water being cooled in cooling<br />
towers should be treated with biocides to check the build up of populations of<br />
undesirable organisms, including beer-spoilage organisms <strong>and</strong> Legionella. For cleaning<br />
in vessels, pipework, bottles, kegs, etc., the water used should be sterile <strong>and</strong> it may<br />
contain traces of sterilant (e.g. ClO2). It must not leave deposits after draining. Water<br />
used for general cleaning, but that does not come into contact with the microbe-free<br />
surfaces that will contact wort or beer, can be of alower quality <strong>and</strong> need not be sterile.<br />
Clearly, supplying liquors of the correct grades for different uses around abrewery can<br />
be acomparatively complex process. The objective must be to obtain supplies of the<br />
various grades as simply <strong>and</strong> inexpensively as possible. The ways in which this is<br />
achieved are very varied.<br />
3.6 The effects of ions on the brewing process<br />
Ions present in brewing water have arange of effects on the production process <strong>and</strong> the<br />
quality of the product (Bak et al., 2001; Comrie, 1967; Moll, 1995; Taylor, 1981,1989).<br />
Inthis section the roles of major ions willbe consideredinturn. It willbe understoodthat<br />
other ions are added to the process stream from the grist <strong>and</strong> from the hops. In addition<br />
solid salts may be added directly to the mash or to the wort. Calcium ions (Ca 2+ ,at. wt.<br />
40.08) serve several important functions in brewing. They stabilize the enzyme -<br />
amylase during mashing <strong>and</strong>, by interacting with phosphate, phytate, peptides <strong>and</strong><br />
proteins in the mash <strong>and</strong> during the copper boil, the pH values of the mash <strong>and</strong> the wort<br />
are usefully reduced. For example:<br />
or<br />
3Ca 2‡ ‡2HPO4 2 ‡2OH !Ca3…PO4† 2 #‡2H2O<br />
3Ca 2‡ ‡2HPO4 2 !Ca3…PO4† 2 #‡2H ‡<br />
If bicarbonate ions are also present (the water has temporary hardness) these can more<br />
than offset the effect of calcium <strong>and</strong> cause a rise in pH (Chapter 4). Perhaps the<br />
concentration of calcium ions should not greatly exceed 100 mg/l in the mashing liquor as<br />
no great advantage is gained from higher doses <strong>and</strong> there is the risk that too much<br />
phosphate may be removed from the wort, <strong>and</strong> the yeast may then have an inadequate<br />
supply. Another recommendation is that calcium should be in the range 20 150 mg/l,<br />
depending on the beer being made.<br />
Calcium oxalate, Ca(COO)2, is deposited as beer stone during fermentations, <strong>and</strong> an<br />
adequate level of calcium ions ensures that the deposition is nearly complete. Crystals of<br />
calcium oxalate formed later in packaged beer provide nuclei for the breakout of carbon<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
dioxide <strong>and</strong> so can cause gushing <strong>and</strong> haze. In mashing the fall in pH caused by calcium<br />
ions favours proteolysis <strong>and</strong> so an increase in FAN, <strong>and</strong> faster saccharification. The more<br />
acid conditions also reduce wort colour, hop utilization <strong>and</strong> favour a reduction in<br />
astringent flavours. Calcium ions favour the formation of agood, flocculent hot break<br />
(trub) <strong>and</strong> yeast flocculation, but they seem to have little effect on flavour.<br />
Magnesium ions (Mg 2+ ,at. wt. 24.32) are needed by many yeast enzymes, such as<br />
pyruvate decarboxylase. In some respects the effects of this ion resemble those of the<br />
calciumion, but the effects onpH from interactions with phosphates are less pronounced,<br />
being about half, because the salts are more soluble. While high concentrations of<br />
magnesium ions are unusual, they can impart asour or bitter flavour to beer. High,<br />
laxative concentrations are not reached. An upper limit of 30mg magnesium ions/litre<br />
has been proposed.<br />
Sodium ions (Na + ,at. wt. 23.0) occur in some waters <strong>and</strong> sodium chloride is the main<br />
solute in saline waters. Sodium ions can impart sour/salty flavours at high concentrations<br />
(over about 150 mg/litre, which is also aproposed maximum concentration) <strong>and</strong> sodium<br />
chloride may be added tobrewing liquors (75 150mg/l) to enhance `palate-fullness' <strong>and</strong><br />
a certain sweetness. Sometimes potassium chloride is added instead, at low<br />
concentrations, to achieve a less sour flavour. Excess potassium ions ((K + , at. wt.<br />
39.1) >10mg/l) can have laxative effects <strong>and</strong> impart asalty taste.<br />
Hydrogen ions (H +, at. wt. 1.01) <strong>and</strong> hydroxyl ions (OH ,at. wt. 17.01) are always<br />
present in water, which is neutral when these ions are present in equimolecular amounts,<br />
[H + ] = [OH ]. The negative log10 of the hydrogen ion concentration, expressed in<br />
molarity, is the pH. As the temperature rises the dissociation of the water increases, the<br />
hydrogenionconcentrationincreases,<strong>and</strong>sothepHofwateratneutrality declines (Table<br />
A8 on page 842).<br />
Iron ions (Fe 2+ , ferrous <strong>and</strong> Fe 3+ , ferric; at. wt. 55.9) can occur in solution, for<br />
example, as ferrous bicarbonate or complexed with organic materials. Ferrous water is<br />
undesirable for brewing purposes, since it can deposit slimes (probably after oxidation, as<br />
red-brown hydrated ferric hydroxide), which can block pipes, filters, ion exchange<br />
columns, reverse osmosis equipment, etc. In addition, iron ions can confer dark colours to<br />
worts <strong>and</strong> beers by interacting with phenolic substances from the malt <strong>and</strong> hops <strong>and</strong> can<br />
convey metallic, astringent tastes to beers, give hazy worts <strong>and</strong> inhibit yeasts. The ions,<br />
possibly because of their ability to act as oxidation/reduction catalysts, favour haze<br />
formation <strong>and</strong> flavour instability. At concentrations of > 1 mg/l iron ions are harmful to<br />
yeasts. Perhaps concentrations should be reduced to less than 0.1 mg Fe/l. For all these<br />
reasons, <strong>and</strong> because of the difficulties that they can cause in some water treatments, it is<br />
usual to reduce the levels of dissolved iron early in a water treatment process.<br />
Copper (Cu 2+ , at. wt. 63.5) presented problems in brewing when vessels <strong>and</strong> pipework<br />
were made of copper but since these have come to be made of stainless steel there have<br />
been fewer problems with dissolved copper in breweries. Copper ions are toxic <strong>and</strong><br />
mutagenic to yeasts, which accumulate them <strong>and</strong> develop `yeast weakness'. Another<br />
source of copper ions was the older, copper-based fungicides applied to hops. Copper<br />
ions are oxidation/reduction catalysts <strong>and</strong> their presence favours flavour instability <strong>and</strong><br />
haze formation in beer. <strong>Brewing</strong> liquor should contain < 0.1 mg copper/litre.<br />
Manganese (Mn 2+ , manganous; Mn 4+ , manganic; at. wt. 54.9) levels in brewing water<br />
should be low, (< 0.2 mg/litre or even < 0.05 mg/litre) but trace amounts of this element<br />
(<strong>and</strong> copper <strong>and</strong> iron) are needed by yeast. Like copper <strong>and</strong> iron these ions are oxidation/<br />
reduction catalysts <strong>and</strong> have adverse effects on flavour <strong>and</strong> beer colloidal stability.<br />
Manganese is less easily removed from water than iron.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Ammonia(NH3,m.wt.17.03)<strong>and</strong>ammoniumions(NH4 + ,m.wt.18.04.pKaˆ9.25at<br />
25ëC) occurs almost entirely as ammonium ions under brewing conditions. Ammonia in<br />
water indicates that the water may be contaminated with rotting organic matter <strong>and</strong> a<br />
proposed maximum concentration is 0.5mg/litre. However, ammonia can escape from<br />
refrigeration equipment <strong>and</strong> this, very water-soluble, gas is toxic.<br />
Zinc (Zn 2+ ,at. wt. 65.4), if present in appreciable amounts in brewing water, usually<br />
inicates thatthision hasbeen pickedupduringtransferorstorage.Highconcentrations in<br />
ground watersare unusual.Athigh levels this substancecan betoxic, the upper permitted<br />
concentration in potable water is 5mg/l (Table 3.2). High concentrations are damaging to<br />
yeasts but small amounts are essential. Not infrequently the levels of zinc in worts are<br />
insufficient to maintain good fermentations <strong>and</strong> in these cases the worts may be<br />
supplemented with additions of zinc chloride (0.15 0.2 mg/l). The recommended range<br />
in brewing liquor is 0.15 0.5 mg/litre.<br />
Bicarbonate (HCO3 , m. wt. 61.02) <strong>and</strong> carbonate ions (CO3 2 , m. wt. 60.01). The<br />
stages of the ionization of carbonic acid, formed by the hydration of carbon dioxide, are:<br />
CO2 ‡ H2O $ H2CO3 $ H ‡ ‡ HCO3 $ 2H ‡ ‡ CO3 2<br />
The pKa values of the first <strong>and</strong> second dissociations, at 25 ëC, are 6.4 <strong>and</strong> 10.3<br />
respectively. Thus at brewing pH values carbon dioxide is present as the gas, as carbonic<br />
acid <strong>and</strong> as the bicarbonate ion. High levels of bicarbonate ions in brewing water are<br />
undesirable since they cause unwanted increases in pH during mashing <strong>and</strong> sparging <strong>and</strong><br />
in the hop-boil. Probably the concentration of bicarbonate ions in brewing liquor should<br />
never exceed 50 mg/l.<br />
Sulphate ions (SO4 2 , m. wt. 96.07); sulphate is the major counter ion to calcium <strong>and</strong><br />
magnesium ions in permanently hard water. The ion contributes a drier, more bitter<br />
flavour to beers that should be balanced by appropriate amounts of chloride ions. Yeasts<br />
metabolize sulphate producing, inter alia, small amounts of hydrogen sulphide, (H 2S),<br />
sulphur dioxide, (SO 2), <strong>and</strong> other substances that contribute to the aromas of beers<br />
brewed with sulphate-rich water. The classic example is the `Burton nose' of the ales<br />
brewed at Burton-upon-Trent. Acceptable sulphate concentrations are in the range<br />
10 250 mg/litre.<br />
Chloride ions (Cl , at. wt. 35.5) occur at high levels in saline waters. High levels are<br />
reported to limit yeast flocculation but to improve beer clarification <strong>and</strong> colloidal<br />
stability. Chloride ions contribute to the mellow, palate-full character of beer. The ratio of<br />
chloride to sulphate helps to regulate the saline/bitter character of beer. Ratios <strong>and</strong><br />
concentrations for different types of beers have been proposed (see above; Comrie, 1967).<br />
A reasonable maximum concentration is 150 mg/litre.<br />
Nitrate (NO3 , m. wt. 62.01) <strong>and</strong> nitrite ions (NO2 , m. wt. 46.01); there is concern<br />
about the rising levels of nitrate ions being found in ground waters. These ions are<br />
derived from agricultural fertilizers being leached from the topsoil <strong>and</strong> filtering down to<br />
the aquifers that supply water. Other sources are sewage <strong>and</strong> rotting organic matter. Even<br />
if the nitrogen is initially present as ammonium ions these are quickly oxidized to nitrate<br />
in the soil. Potable water usually has an upper limit of 50 mg nitrate/l <strong>and</strong> a limit of 0.1 or<br />
0.5 mg nitrite/l, <strong>and</strong> these limits must not be exceeded in beers. However, brewers require<br />
lower levels in their brewing water since nitrate will be added to wort from the hops. The<br />
concern is that bacterial contaminants may reduce nitrate to nitrite. Nitrite ions can be<br />
toxic, they can give colours with tannins <strong>and</strong>, of most concern, can give rise to potentially<br />
carcinogenic nitrosamines.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Phosphate ions (e.g. HPO3 2 , m. wt. 80.01); the stages of ionization of phosphoric<br />
acid are:<br />
H3PO4 $ H ‡ ‡ H2PO4 $ 2H ‡ ‡ HPO4 2 $ 3H ‡ ‡ PO4 3<br />
The pK a values for the successive dissociations, at 25 ëC, are 2.0 2.2, 6.7 6.8 <strong>and</strong> 12.4.<br />
Thus at mashing, wort <strong>and</strong> beer pH values, around pH5, most phosphate is present as the<br />
H2PO4 ion. There are regulatory limits on the concentration of phosphate phosphorus<br />
that may be present in potable water, <strong>and</strong> a suggested maximum in brewing liquor is 1<br />
mg/litre. Most of the phosphate in beer is derived from malt, although phosphoric acid or<br />
acid phosphate salts may be used to adjust the pH or to release carbon dioxide from<br />
bicarbonate-rich waters. Worts can also contain the organic phosphate phytate (salts of<br />
phytic acid), derived from malt. Phosphates are important pH buffers in brewing <strong>and</strong><br />
interactions between calcium ions <strong>and</strong> phosphates, <strong>and</strong> other substances, usefully reduce<br />
the pH in mashing <strong>and</strong> during the hop-boil. Phosphoric acid is also used for acid-washing<br />
yeasts.<br />
Silicate ions (e.g., SiO3 2 , , m. wt. 76.09); silica can dissolve to form a range of ions,<br />
with various silica to oxygen ratios. Reportedly high concentrations of silicates can<br />
damage yeasts <strong>and</strong> give rise to hazes in beers but if so, these events must be infrequent.<br />
The most significant effect of silicates is the deposition of scales, formed between silicate<br />
<strong>and</strong> calcium <strong>and</strong> magnesium ions, when the water is heated. An upper acceptable<br />
concentration of silicate may be 40 mg/litre.<br />
Fluoride ions (F , at. wt. 19.00) occur in some ground waters. They rarely reach toxic<br />
levels. Potable waters have upper concentration limits of about 1.5 mg/litre. Small<br />
amounts of fluorides may be added to water for domestic use to reduce the incidence of<br />
dental caries. Even at substantially higher levels (10 mg F/litre) the ions are without<br />
perceptible effects on brewing.<br />
3.7 Brewery effluents, wastes <strong>and</strong> by-products<br />
Breweries generate wastes, by-products, pollutants <strong>and</strong> effluents (Armitt, 1981; Brooks et<br />
al., 1972; Huige, 1994; Isaac, 1976; Isaac <strong>and</strong> Anderson, 1973; Klijnhout <strong>and</strong> Van Eerde,<br />
1986; Meyer, 1973; Rostron, 1996). These must be dealt with in the least costly way or,<br />
in one or two instances, profitably. Into these categories come noise, heat, odours, dusts<br />
(from malt <strong>and</strong> adjuncts), cullet (broken glass), waste aluminium cans, plastic waste,<br />
domestic <strong>and</strong> laboratory wastes, carbon dioxide, trub, spent grains, spent grain drainings<br />
<strong>and</strong> pressings, surplus yeast, used kieselguhr from filters, waste beer, wort, waste water,<br />
boiler blow-down sludge, <strong>and</strong> acids, alkalis <strong>and</strong> detergents (from CIP <strong>and</strong> other cleaning<br />
systems), labels, lubricants, copper condensate, PVPP <strong>and</strong> ion exchange regeneration<br />
reagents, <strong>and</strong> so on. Dealing with these is expensive. Breweries generate large volumes of<br />
waste water, <strong>and</strong> most of the remainder of this chapter is concerned with its disposal.<br />
However, all wastes <strong>and</strong> by-products must be disposed of quickly in the interest of saving<br />
space <strong>and</strong> minimizing the risks of microbial contamination. Broadly, all wastes are either<br />
dumped or disposed of for recycling or sold or discharged into a sewer or a waterway or<br />
the sea. Once it was normal to wash as much waste as possible down a sewer. In many<br />
countries this is now too costly <strong>and</strong> so attempts are made to recycle materials, to<br />
minimize waste production, to `add value' to by-products or at least to sell them <strong>and</strong> to<br />
partly or completely treat liquid wastes.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
3.7.1 The characterization of waste water<br />
Brewery waste water is chiefly contaminated with putrescible organic matter <strong>and</strong> so it is<br />
comparatively easily purified by biological treatments. However, the flow of waste water<br />
from a brewery varies greatly with the time of day, often with the day of the week <strong>and</strong><br />
with the time of year. Worse, the water will vary widely in its temperature, pH, load of<br />
suspended solids <strong>and</strong> the amounts of organic <strong>and</strong> inorganic materials in solution. If the<br />
waste is being discharged to a public sewer the operating authority will usually set limits<br />
on the composition, volume, rate of flow, temperature <strong>and</strong> pH of the effluent, with<br />
swingeing penalties if the limits are exceeded. Evidently the detailed composition of the<br />
water is variable <strong>and</strong> very complex. However, for treatment purposes water analyses are<br />
simplified, <strong>and</strong> these analyses are adequate for charging for treatments <strong>and</strong> for deciding<br />
what the correct treatments should be (Armitt,1981; Benson et al., 1997; Briggs et al.,<br />
1981).<br />
Suspended, or settleable, solids, SS, are usually reported as mg dry matter/litre.<br />
Sometimes SS are defined as particles retained by a 1 m filter. Their presence results in<br />
a reduced water clarity <strong>and</strong> they may deposit <strong>and</strong> create blockages or be abrasive <strong>and</strong><br />
cause wear to pumps. Some may be partly destroyed or removed during biological<br />
treatments. Suspended solids which are not biodegradable usually finish as sludge in<br />
treatment plants. Total dissolved solids, TDS mg/litre, dissolved organic carbon, DOC<br />
mg/litre, <strong>and</strong> total organic carbon, TOC mg/litre are measures sometimes used but much<br />
more emphasis is placed on the biological oxygen dem<strong>and</strong>, BOD, <strong>and</strong> the chemical<br />
oxygen dem<strong>and</strong>, COD.<br />
Because most older waste-water treatments were based on microbiological aerobic<br />
oxidations of dissolved organic matter, the biological, or biochemical, oxygen dem<strong>and</strong> of<br />
the waste was of prime importance. The BOD5 20 is the weight of oxygen (mg) taken up as<br />
the organic substances in the water (1 litre) are oxidized by a mixture of micro-organisms<br />
in five days at 20 ëC, in the dark. The test must be in darkness to prevent the growth of<br />
algae, which generate oxygen. This important test is slow <strong>and</strong> not very precise. In<br />
consequence it is increasingly being replaced with determinations of the chemical oxygen<br />
dem<strong>and</strong>, COD. This is normally calculated from the amount of dichromate used up when<br />
the water is boiled with the acid reagent for two hours in the presence of a silver sulphate<br />
catalyst. The results are expressed as mg oxygen/litre water. However, as the dichromate<br />
oxidizes more substances than the microbes the COD values are greater than the BOD<br />
values. The ratios between these values vary widely, but for most mixed brewery wastes<br />
the COD/BOD ratio is 1.6 1.8. For domestic sewage the value is about 2.5. The<br />
permanganate value, PV, an alternative determination of oxidizable organic matter, is<br />
now used less.<br />
In the UK the costs of having effluent treated in a municipal sewage works is usually<br />
calculated with reference to the `Mogden Formula', (named after a London sewage<br />
works) or a modification of it.<br />
C ˆ R ‡ V ‡ …OT=OSxB† ‡ …ST=SSxS†<br />
Where C ˆ the total charge/m 3 (unit volume) for the trade effluent discharge. R ˆ the<br />
cost of conveying <strong>and</strong> receiving the effluent + overhead costs. V ˆ unit cost of<br />
volumetric <strong>and</strong> primary treatments (screening <strong>and</strong> settlement). O T ˆ COD of the trade<br />
effluent after settlement at pH7. OS ˆ COD of the average settled sewage <strong>and</strong> trade<br />
effluent (i.e. a reference strength). B ˆ the unit cost of the biological treatment of the<br />
mixed <strong>and</strong> settled trade effluent <strong>and</strong> sewage. ST ˆ total suspended solids (SS) of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mixed trade effluent. SSˆThe SS of the mixture of the sewage <strong>and</strong> trade effluent,<br />
(st<strong>and</strong>ard strength). Sˆthe unit cost of the treatment <strong>and</strong> disposal, of the sludge. Thus<br />
increases of COD, SS or volume lead to greater charges.<br />
Extracharges willbe madefor effluents discharged outside stipulated`consent limits',<br />
such as peak volume flows, pH or temperature values, COD, SS, soluble nitrogen, <strong>and</strong> so<br />
on. The actual charges made vary from place to place <strong>and</strong> are continuing to rise,<br />
pressurizing brewers to find ways of reducing these costs. Using particular examples,<br />
with 1987 figures, Table 3.3 illustrates how increasing water usage, <strong>and</strong> the inevitable<br />
extra effluent production, can be very expensive. Examples of reported ranges of BOD<br />
(mg/litre) are beers, 60,000 120,000; spent grain press liquor, 60,000; waste yeast with<br />
entrained beer, 200,000; trub, 70,000; spent kieselguhr, 80,000; fermenter washings,<br />
20,000 30,000. Suspended solids, SS, represent 12 23% of brewery effluent BOD<br />
(Armitt, 1981; Benson et al., 1997). Examples of the average characteristics of brewery<br />
waste water are: BOD5(mg/litre), 400 1750 (900 2000); mean COD(mg/litre),<br />
1200 3,000; BOD5(kg/hl product), 0.45 0.95; SS(mg/litre), 93 772; SS (kg/hl<br />
product), 0.17 0.40; pH4.1 11.5; temperature 13 49ëC (55.4 120.2ëF), ratio of<br />
effluent volume produced/volume of beer produced, 4 33. Soluble nitrogen (mg/litre),<br />
30 80; phosphorus (mg/litre), 10 30. However, shock discharges, in particular effluent<br />
flows may be pH 2 3to 10 13 <strong>and</strong> BOD 170,000 500,000. These should be prevented<br />
from reaching the sewer (see below). Abrewery making 10,000 30,000 barrels of beer/<br />
week may release in the effluent 2 5.6tCOD/day <strong>and</strong> 0.46 0.87tSS/day.<br />
Sometimes brewery effluent loads are expressed as population equivalents. Different<br />
equivalentsmaybeused(Armitt,1981).However,abrewerywithanoutputof10 6 hlwill<br />
produce about as much effluent as 50,000 people. The quality of treated water that may<br />
be discharged depends on whether this is to the sea, to an estuary, or to ariver, <strong>and</strong> what<br />
the minimum dilution rate will be. Some suggested limits are COD, 127mg/litre; BOD,<br />
25mg/litre; SS, 35mg/litre; nitrogen, 10mg/litre; phosphorus, 1mg/litre. Exceeding<br />
these, or other appropriate limits, will risk causing algal blooms, or the overgrowth of<br />
Table 3.3 Examples of calculated water <strong>and</strong> effluent costs, using mean UK Water Authority<br />
charges for 1987 (data of Askew, 1987). A brewery has an annual production of 1 10 6 barrels<br />
(36 10 6 imp. gal.; 1.637 10 6 hl). With a water : product ratio of 8 : 1, an effluent product ratio of<br />
5.5 : 1, a mean effluent BOD of 1000 mg/l, a mean COD of 1800 mg/l, <strong>and</strong> an effluent mean SS of<br />
500 mg/l then the water dem<strong>and</strong> p.a. is 288 10 6 imp. gal., the effluent volume is 198 10 6 imp.<br />
gal, the BOD load p.a. is 900 tonnes (t), the COD load p.a. is 1620 t/p.a. <strong>and</strong> the SS load is 450 t/p.a.<br />
Assuming different efficiencies of water usage the three examples show the variations in water <strong>and</strong><br />
effluent costs<br />
Measure Case 1 Case 2 Case 3<br />
Water: product ratio 8 : 1 9.23 : 1 12.9 : 1<br />
Annual water consumption (m 3 ) 1.309 10 6<br />
1.51 10 6<br />
2.11 10 6<br />
Cost (p/m 3 ) 27.8 27.8 27.8<br />
Cost (£/year) 363,902 419,780 586,580<br />
Effluent: product ratio 5.5 6.43 9.7<br />
Effluent volume (m 3 ) 900,000 1,052,180 1,587,270<br />
Effluent COD (mg/l) 1,800 3,580 4,500<br />
Effluent SS (mg/l) 500 291 400<br />
Effluent cost (p/m 3 ) 43.52 62.85 77.95<br />
Effluent cost (£/year) 391,680 661,295 1,237,277<br />
Total water <strong>and</strong> effluent costs (£/year) 755,582 1,081,075 1,823,857<br />
Since 1987 costs have risen to a considerable extent. At present (August, 2004) £1 ˆ 1.48 Euros ˆ $ (USA) 1.80<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
microbescausing the receiving water tobecome anaerobicsokillingallhigherlifeforms.<br />
Clearly brewery wastes must be treated before discharge to waterways. The treatment<br />
may be carried out, in whole or in part, at the brewery or at asewage works.<br />
3.7.2 The characteristics of some brewery wastes <strong>and</strong> by-products<br />
The immense range of ratios of water taken in to beer produced is chiefly due to different<br />
efficiencies of water use, although some breweries, e.g., those that bottle a high<br />
proportion of their beer in returnable bottles, (which must be cleaned), are at a<br />
disadvantage.Tocontrolwasteitisnecessarytometerthevolume<strong>and</strong>compositionofthe<br />
effluent from every department <strong>and</strong> the brewery as a whole, to detect <strong>and</strong> prevent<br />
wasteful <strong>practice</strong>s. To be of use effluent streams must be sampled in statistically valid<br />
ways. Time-proportional or volume-proportional sampling may be used. Apparently<br />
minor leaks, or leaving hoses running, or having CIP programmes operating with<br />
excessive water rinsing or without the re-use of final rinse liquor for the first rinse can<br />
cause substantial losses (Horrigan et al., 1989). The pH of effluent may become extreme,<br />
for example, when alkaline cleaning liquids or PVPP or ion exchange regeneration<br />
liquors are released. Precautions should be taken to trap these liquors <strong>and</strong> release them<br />
slowly at ametered rate with the general effluent to dilute them to an acceptable level, or<br />
their pH values may have to be adjusted before release. It has been proposed that alkaline<br />
liquors should be neutralized with carbon dioxide from furnace gases or from the<br />
fermenters rather than with mineral acids.<br />
Much effort has been spent in finding better ways of dealing with brewery by-products<br />
<strong>and</strong> wastes (Brooks et al., 1972; Horrigan et al., 1989; Huige, 1994; Penrose, 1985; Reed<br />
<strong>and</strong> Henderson,1999/2000; Vriens et al., 1986, 1990). Many breweries collect some of the<br />
carbon dioxide from the fermenters <strong>and</strong> use it to carbonate beer (Chapter 15). Others allow<br />
much of it to escape <strong>and</strong> yet others may use it to neutralize alkaline waste liquors. Furnace<br />
gases, which may also be used for neutralizing alkaline effluents, contain carbon dioxide<br />
<strong>and</strong> acid oxides of sulphur. Efforts are now made to save heat in breweries (re-use of<br />
cooling water, warming water by condensing vapours from the hop-boil, etc.), <strong>and</strong> so less<br />
heat escapes in effluents. In principle, heat could be withdrawn from effluents using heat<br />
pumps, but probably this is not economic. It has been suggested that warm, clean water<br />
from a brewery might be used in a fish farm. Trub (hot break) <strong>and</strong> spent hops usually<br />
contain some wort. If washed into the sewer they add substantially to the BOD <strong>and</strong> SS.<br />
Normally the entrained wort is recovered, for example, by adding the trub <strong>and</strong> spent hops to<br />
the lauter tun. Alternatively, the spent hops <strong>and</strong> trub may be added directly to the spent<br />
grains. Some 30% of the trub solids are digestible proteins that add to the feed value of the<br />
spent grains. Spent hops have been used as a mulch or as low-grade fertilizer.<br />
Sometimes trub is mixed with surplus yeast intended for animal feed. Surplus yeast is<br />
collected from fermenters, <strong>and</strong> is also present in tank bottoms. It can add substantially to<br />
the BOD <strong>and</strong> SS of effluents <strong>and</strong>, with a crude protein content of about 47% d.m., a<br />
carbohydrate content of 43% d.m., <strong>and</strong> a mixture of vitamins surplus yeast is potentially a<br />
valuable by-product. Often it is sold to companies that debitter it <strong>and</strong> turn it into food<br />
supplements (Putman, 2001). Surplus brewing yeast is used by distillers. On a<br />
comparatively small scale the yeast may be used as a source of biochemicals, particular<br />
proteins, enzymes <strong>and</strong> glutathione, <strong>and</strong> yeast extracts are used in culture media for<br />
microbes. Yeast tablets are used as vitamin supplements. Sometimes the yeast cake is<br />
washed <strong>and</strong> the yeast is returned to the production stream by adding it into the mash.<br />
Alternatively, it may be autolysed <strong>and</strong> then added to the spent grains to enhance their<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
valueasanimalfeed.Maltdusthasbeenaddedtomashesorhasbeenmixedinwithspent<br />
grains.<br />
Used kieselguhr, from beer filters, used to be flushed into the sewers, but its large<br />
contribution to SS <strong>and</strong> BODs makes this undesirable <strong>and</strong> probably most is dumped, with<br />
or without first mixing it with quick-lime, at l<strong>and</strong>fill sites. Small amounts have been used<br />
as asoil improver. Sometimes the used filter-aid has been mixed with the spent grains,<br />
but this is not always acceptable. Dumping is becoming costly <strong>and</strong> other disposal<br />
methods are being tested. Filter aid has been regenerated by chemical cleaning <strong>and</strong> by<br />
calcining to burn away organic residues. The regenerated material has been used again as<br />
afilter aid, as afertilizer carrier, <strong>and</strong> as afiller in paints <strong>and</strong> varnishes.<br />
`Waste' beer or wort may include last runnings from the mash, press-liquor from the<br />
spent grains, rinsingsfrom vessels orpipework,returnedbeer, spillages from bottling<strong>and</strong><br />
canning plants <strong>and</strong> rinsings from returned containers. All these residues may<br />
(expensively) be consigned to the sewer. Some brewers believe that returning last<br />
runnings <strong>and</strong> liquor from the spent grains to the mash constitutes arisk to the quality of<br />
the beer. Others have used these materials after clarification by centrifugation, with or<br />
without treatment with active charcoal, <strong>and</strong> keeping them for not more than two to three<br />
hours at 80ëC before adding them to the mashing liquor or to the copper (Coors <strong>and</strong><br />
Jangaard, 1975). In these instances no deterioration of beer quality was detected.<br />
Returned beer may be blended or discarded. Returned beer <strong>and</strong> weak worts have been<br />
used to make alcohol or vinegar, to grow other microbes, such as fodder yeasts (Torula<br />
spp., C<strong>and</strong>ida spp., Aspergillus spp.), for use in foodstuffs. Yet others have dried the<br />
liquids. The product, dried brewers' solubles, has been added to foodstuffs. Probably<br />
most of these processes are not usually economically viable.<br />
Spent grains are produced at the end of every mash. They are of value as afoodstuff,<br />
particularly for ruminants, but they are bulky, <strong>and</strong> they soon begin to decompose, so they<br />
mustberemovedfromthebrewerypromptly.Theh<strong>and</strong>lingequipmentmustbekeptclean<br />
to prevent the growth of spoilage organisms. Depending on the grists their composition is<br />
variable. In one example the composition reported, on adry weight basis, was crude<br />
protein, 27%; fat, 6 7%; ash, 4 5%; crude fibre, 15%; N-free extract, 46%. The<br />
moisture content of spent grains varies widely depending on the wort separation system<br />
used (Chapter 6). Thus grains from aStrainmaster may contain 87 90% moisture, <strong>and</strong><br />
they are sloppy <strong>and</strong> are easy to pump, but they are so wet that liquid drains from them <strong>and</strong><br />
they must be de-watered before removal. Water drains from grains with moisture contents<br />
above 80%. The drainings are an excellent medium for unwanted microbes. Grains from<br />
lauter tuns can contain 75 85% moisture <strong>and</strong> those from pressure filters contain as little<br />
as 50 55%. Exceptionally, these grains may be dried further in a current of hot air, to c.<br />
8% moisture, when the dried material is stable. Grains may be sold for animal feed either<br />
wet or after de-watering. Using continuous screw or roller presses the moisture contents<br />
can be reduced to 63 72%, producing a more easily h<strong>and</strong>led solid material <strong>and</strong> the<br />
squeeze liquor. The liquor may contain up to 3.5% dissolved solids <strong>and</strong> 5% SS as it has a<br />
high BOD. As it contains valuable extract it should be returned to the process stream,<br />
where possible. In one case this saved about 1% of brewer's extract <strong>and</strong> reduced water<br />
use by 5% (Coors <strong>and</strong> Jangaard, 1975).<br />
The grains used for animal feed must be stored by farmers <strong>and</strong> the drier they are the<br />
easier they are to h<strong>and</strong>le. Sometimes they are ensiled <strong>and</strong> additions of propionic acid or<br />
sorbic acid with phosphoric acid have been used to act as preservatives, <strong>and</strong> seeding with<br />
lactic acid bacteria has been proposed. As noted, autolysed yeast, trub, spent hops <strong>and</strong><br />
even used kieselguhr may be added to spent grains. A range of other uses for spent grains<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
has been proposed, as asource of biogas <strong>and</strong> soil conditioner produced by anaerobic<br />
digestion, disposal by burning (giving heat), as asource of `secondary worts' generated<br />
by acid orenzymic hydrolysis, as asourceof protein, as asourceof food-gradefibre,asa<br />
basis for mushroom compost, as asoil conditioner <strong>and</strong> organic fertilizer, as amedium for<br />
growing earthworms to use in poultry food, <strong>and</strong> in fish food. None of these alternative<br />
uses seems to be widely employed.<br />
3.8 The disposal of brewery effluents<br />
Aconsideration of the sources of brewhouse effluents shows that large volumes, BOD<br />
<strong>and</strong> SS are generated at different stages of the brewing process (Table 3.4; Askew, 1975,<br />
1987; Askew <strong>and</strong> Rogers, 1997; Huige, 1994; Love, 1987; Robertson et al., 1979). In<br />
some breweries all the `waste' water is collected into a common sewer. However, it is<br />
senseless for surface run-off (storm water) to be directed to a treatment plant <strong>and</strong> clean,<br />
but warm water that has been used for cooling should find a use in the brewhouse for<br />
mashing or cleaning, both to minimize water <strong>and</strong> effluent charges <strong>and</strong> to conserve heat<br />
<strong>and</strong> so reduce heating costs. If the effluents are treated, in whole or in part, at the brewery<br />
then it is usually advisable to separate `weak' <strong>and</strong> `strong' effluents <strong>and</strong> to treat them<br />
separately. The costs of treating effluents in municipal sewage treatment works are high.<br />
Brewers try to minimize the volumes <strong>and</strong> strengths (BOD, COD, SS) of the effluents.<br />
Sometimes it is economical to partly, or even extensively, treat brewery effluents on site.<br />
Secondary treatment may allow the water to be discharged into a watercourse. If costs<br />
continue to rise it may become worthwhile to purify some effluents further, using tertiary<br />
treatments, so that the water is pure enough to be used again for some purposes in the<br />
brewhouse <strong>and</strong> so, by recycling, avoid acquisition <strong>and</strong> disposal costs. Usually effluent<br />
purification is undertaken with reluctance. Treatment plant takes up space, it is costly,<br />
<strong>and</strong> it requires well-trained staff to operate it successfully.<br />
Whether effluents are treated at a brewery site or elsewhere the objectives are the<br />
same, to reduce the temperature to a moderate level (often under 40 ëC, 104 ëF), to restrict<br />
the pH to a specified range (e.g. 6 10) <strong>and</strong> to reduce the BOD, COD <strong>and</strong> SS levels to<br />
below specified levels (e.g. 25, 125 <strong>and</strong> 35 mg/l, respectively) so that the water can be<br />
released into a stream, river or estuary. Some breweries do not carry out any treatment on<br />
site, <strong>and</strong> there is no uniform system of treatment among the others. Historically, after<br />
preliminary screening, water was purified, using oxidative, aerobic biological systems. In<br />
recent years partial treatments using some anaerobic processes are being used. It is<br />
convenient to divide treatments into preliminary treatments, `primary' treatments,<br />
`secondary' treatments (aerobic, anaerobic or a combination of the two), <strong>and</strong> `tertiary' or<br />
`polishing treatments'.<br />
3.8.1 Preliminary treatments<br />
Many preliminary treatments are in use (Armitt, 1981; Benson et al., 1997; Huige, 1994;<br />
Klijnhout <strong>and</strong> Van Eerde, 1986; Vereijken et al., 1999; Vriens et al., 1986, 1990; Walker,<br />
1994). First the effluent should be screened to remove labels, bottle caps, floating plastic<br />
items <strong>and</strong> spent grains. These screens may be of many types, for instance, hyperbolic bar<br />
screens or screens of woven stainless steel mesh. It is desirable that the water also flows<br />
through a settling tank in which the reduction of the flow-rate permits cullet, grit, s<strong>and</strong><br />
<strong>and</strong> some SS to settle. The deposit formed is removed at intervals by scrapers, to be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 3.4 Estimated biological oxygen dem<strong>and</strong> <strong>and</strong> suspended solids loads in a brewery employing yeast recovery (Robertson et al., 1979)<br />
Process Flow BOD SS<br />
(hl/day) (gal/day) (kg/day) (lb/day) (kg/day) (lb/day)<br />
Mash mixer 69.4 1526 3.6 8 7.3 16<br />
Lauter tun* 68.0 1496 189.6 418 122.5 270<br />
Brew kettle (copper) 97.2 2137 27.7 61 3.6 8<br />
Hot wort tank (whirlpool)* 26.6 584 284.0 626 98.0 216<br />
Wort cooler 24.1 531 0.5 1 0 0<br />
Fermenters* 290.0 6386 240.4 530 155.1 342<br />
Ageing tank (maturation) 407.7 8962 93.0 205 137.9 304<br />
Primary filter* 81.9 1801 46.3 102 231.3 510<br />
Secondary storage 385.1 8471 40.4 89 59.0 130<br />
Secondary filter* 92.6 2037 8.2 18 42.2 93<br />
Bottling tank 39.9 877 0.5 1 0 0<br />
Filler 922.4 20291 23.1 51 9.5 27<br />
Pasteurizer 3582.6 78808 34.5 76 3.2 7<br />
Bottle washer* 2933.5 64529 68.0 150 29.0 64<br />
Cooling water 4332.2 95298 0 0 0 0<br />
Miscellaneous flows 92.2 2028 0.5 1 0.5 1<br />
*Major sources of BOD <strong>and</strong>/or SS. Gal ˆ imperial gallons. (1 imp. gal ˆ 1.201 US gal ˆ 4.546 litres).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
dumped. Suspended kieselguhr <strong>and</strong> some yeast may be removed at this stage. The<br />
effluent may also pass through oil traps. The compositions, characteristics <strong>and</strong> flows of<br />
brewery effluents are highly variable <strong>and</strong> need to be `evened out' if treatments are to be<br />
successful. Strongly acid or alkaline cleaning or regeneration solutions are sometimes<br />
released immediately into the effluent stream. It is better <strong>practice</strong> to hold these solutions<br />
in a special receiver to mix acidic <strong>and</strong> alkaline solutions <strong>and</strong>, perhaps after adjusting the<br />
pH, release them into the main flow of effluent over a period of time to dilute them to<br />
such an extent that they are harmless. Sometimes such material may be removed by a<br />
specialist contractor. An additional `calamity tank' may be provided to hold sudden<br />
unexpected flows of liquid that can be released later, over a period, <strong>and</strong> so even out<br />
variations in the composition <strong>and</strong> flow rate of the effluent. All breweries, whether or not<br />
they treat their own effluents, should have a balancing or conditioning tank. This should<br />
be stirred <strong>and</strong> possibly be aerated to prevent the formation of odours. The size should be<br />
decided with reference to the brewery operations <strong>and</strong> should have a residence time (12<br />
hours <strong>and</strong> 36 hours have been suggested, as well as much longer periods) selected to truly<br />
`average out' effluent flow rate <strong>and</strong> composition, so that it will not be harmful to the<br />
microbes that carry out the next stage of treatment. The pH of the effluent may be<br />
adjusted automatically as it leaves this tank <strong>and</strong> if it is going to a biological treatment<br />
plant on site some microbial nutrients may be added at this stage.<br />
Sometimes a different type of preliminary treatment may be carried out. For example,<br />
the effluent may be treated with a slurry of lime or another inorganic coagulant, <strong>and</strong><br />
perhaps a synthetic polyelectrolyte. Then the mixture is given a flotation treatment. Air is<br />
dissolved in the effluent under pressure <strong>and</strong> this is directed into the base of the flotation<br />
vessel. The air separates as a cloud of fine bubbles <strong>and</strong> carries the coagulated materials up<br />
to the top of the vessel from which they are skimmed. In one case the removal of the SS<br />
was 96% <strong>and</strong> of the COD, 6% while in another instance the values were 60% <strong>and</strong> 45%<br />
respectively (Hughes, 1987; Lunney, 1981).<br />
3.8.2 Aerobic treatments of brewery effluents<br />
Many aerobic systems have been tried for treating brewery effluents <strong>and</strong> these are well<br />
understood (Armitt, 1981; Benson et al., 1997; Klijnhout <strong>and</strong> Van Eerde, 1986;<br />
KuÈhtreiber <strong>and</strong> Laa-Thaya, 1995; Reed <strong>and</strong> Henderson, 1999/2000). The effluent,<br />
preferably of uniform composition <strong>and</strong> having the correct pH <strong>and</strong> levels of supplementary<br />
nutrients (nitrogen <strong>and</strong> phosphate), is aerated in the presence of `sludge', a mixed<br />
population of micro-organisms. These multiply <strong>and</strong> grow. About 30% of the BODgenerating<br />
substances is oxidized to carbon dioxide <strong>and</strong> water, while the remainder is<br />
assimilated into microbial mass, the sludge. Surplus sludge is collected <strong>and</strong> must be<br />
disposed of. Sludge treatment <strong>and</strong> disposal is often the most difficult part of aerobic<br />
treatments. Broadly aerobic systems are operated in two different ways, although<br />
intermediate methods may be used. In the `low load' methods, effluent `lightly' loaded<br />
with BOD is supplied to a well-aerated mass of microbes <strong>and</strong> the contact time is<br />
extended. Under these conditions a type of sludge develops that is easy to h<strong>and</strong>le <strong>and</strong><br />
settles readily. BOD removal may exceed 98%. In `high load' systems the ratio of BOD<br />
supplied/unit of biomass is high. BOD removal is less, for example 80%. A large mass of<br />
`putrescible sludge' is formed. With effluents having high BOD values `sludge bulking'<br />
occurs, involving the excessive growth of filamentous microbes that do not readily settle.<br />
This sludge is difficult to h<strong>and</strong>le <strong>and</strong> de-water. This system is less resistant to `shock<br />
loads' than the low load system.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
From these considerations it can be concluded that for optimal treatments brewery<br />
effluents should pass through a buffering tank (or two tanks if the high BOD <strong>and</strong> low<br />
BOD effluents are to be treated separately), <strong>and</strong> effluents with high BODs should receive<br />
a preliminary treatment (aerobic or anaerobic; see below) before they are treated further<br />
in a `low-load' system. Aerobic treatment systems are considered in two groups, the<br />
activated sludge systems in which the biomass is in suspension <strong>and</strong> systems in which at<br />
least most of the biomass is attached to a solid support.<br />
Perhaps the `single tank systems' are the most simple of those using activated sludge.<br />
In these effluent is progressively loaded into an aerated tank which already contains<br />
biomass which is mixed by the aeration process, usually achieved by blowing compressed<br />
air into the base of the tank through dispersers which release it as fine bubbles. When the<br />
tank is full the effluent is diverted to the next tank in line. Aeration is continued in the<br />
first tank until the BOD has been sufficiently reduced. Then aeration ceases <strong>and</strong> the<br />
active sludge is allowed to settle <strong>and</strong> the clear, treated effluent is drawn off from above<br />
the sludge. Any surplus sludge is removed, usually to a settling tank, <strong>and</strong> aeration is<br />
resumed <strong>and</strong> the tank is ready to begin receiving the next charge of effluent.<br />
Where large areas of ground are available effluents may be treated in large lagoons.<br />
Usually two lagoons operate in series <strong>and</strong>, because they are large, they constitute a lowload<br />
system. The lagoons are aerated, preferably with bubbles rising from the bottom,<br />
(surface aerators are inefficient <strong>and</strong> can create microbe-laden aerosols) <strong>and</strong> each may be<br />
followed by a sludge-settling tank. Effluent then flows into a lagoon, is diluted <strong>and</strong> aerated.<br />
After an average holding period, it flows into a settling tank. The clarified effluent flows to<br />
the next lagoon or out of the system while a proportion of the sludge is transferred back to<br />
the inlet at the entrance of the lagoon <strong>and</strong> is mixed with the incoming effluent, creating an<br />
initial high sludge concentration. With this <strong>and</strong> other aerobic systems, provided that<br />
aeration is adequate, the higher the sludge concentration the faster the effluent can be<br />
treated. In one case it was reported that BOD removal was about 95% in the first lagoon<br />
<strong>and</strong> increased to 99% in the second. The comparable COD values were 94 <strong>and</strong> 98%.<br />
Numerous more compact activated sludge systems, in which effluent flows through a<br />
series of aerated tanks <strong>and</strong> settling tanks, have been described. Several designs have been<br />
used in breweries. The Pasveer ditch consists of a shallow continuous ditch, roughly<br />
elliptical in plan, <strong>and</strong> with a trapezoidal cross-section, into which effluent flows <strong>and</strong> from<br />
which it flows to a settling tank. Some of the settled sludge is returned to the effluent<br />
inlet. The liquid is kept aerated <strong>and</strong> flowing around the ditch by brush beaters. The<br />
average retention time of effluent is often 2 3 days. When overloaded with effluent there<br />
are problems with sludge bulking, the plants occupy a large area <strong>and</strong> are in the open.<br />
Often breweries must use the least ground possible <strong>and</strong> the plant must not be obtrusive,<br />
particularly if situated in a town. Tall, thin, fully enclosed aerobic reactors may be<br />
chosen. An example is the deep-shaft reactor. These are typically 0.5 2.0 m<br />
(1.64 6.56 ft.) in diameter <strong>and</strong> extend 100 200 m (329 658 ft.) into the ground. In<br />
the centre of the shaft there is a downflow pipe <strong>and</strong> the annular space between the pipes<br />
acts as the riser. Initially the internal circulation is started by filling the shaft with effluent<br />
primed with activated sludge <strong>and</strong> then injecting air into the outer shaft. When the flow is<br />
established air is injected into the inner, down-shaft <strong>and</strong> as the bubbles are carried<br />
downwards it is efficiently dissolved as the pressure rises. As the pressure declines during<br />
the rise in the outer, annular shaft air <strong>and</strong> carbon dioxide break out of solution <strong>and</strong> the<br />
bubbles act as a gas-lift <strong>and</strong> maintain the circulation.<br />
High removals of BOD from brewery effluents have been obtained (Lom <strong>and</strong><br />
Fedderson, 1981; Vriens et al., 1990). A comparable, above-ground tower system is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
described later (Fig. 3.2). One type of atwo-stage aerobic system is the Artois Unitank<br />
(Eyben et al., 1985; Vriens et al., 1986, 1990). In this system the effluent flows through a<br />
rectangular equalization tank, then through two rectangular aerobic/sludge settlement<br />
tanks working in series <strong>and</strong> so comprising a high-loaded tank <strong>and</strong> a low-loaded tank.<br />
Each aerobic/settlement tank is incompletely partitioned into three sections, so that liquid<br />
can move from one section into the next, but free mixing between the sections is<br />
prevented. The effluent flows from the equalization tank into the first compartment of the<br />
high-loaded treatment tank. It moves from compartment to compartment, meeting<br />
aeration <strong>and</strong> activated sludge in the first two. In the third compartment there is no<br />
aeration, the sludge settles <strong>and</strong> the treated effluent leaves, over a sludge-retaining weir,<br />
<strong>and</strong> flows to the low-loaded tank. At intervals of about two hours the flow in a tank is<br />
reversed <strong>and</strong> aeration now occurs in what was previously the settlement compartment<br />
(which is rich in settled sludge) <strong>and</strong> in the central compartment, but what was previously<br />
the inlet compartment is no longer aerated <strong>and</strong> becomes the settlement compartment. The<br />
third, low-loaded tank is operated in a similar way. BOD removal in the first tank is<br />
reported to be 80 88%, <strong>and</strong> to exceed 98% after the second tank treatment. At intervals<br />
surplus sludge is removed to a thickening tank. Promising trials with a novel, pilot-scale<br />
membrane reactor have been reported (Ward, 2000). The effluent from the enclosed,<br />
aerated <strong>and</strong> stirred chamber passed out by way of a membrane filter that retained all the<br />
suspended solids <strong>and</strong> microbes. Ferric sulphate, which acted as a coagulant, <strong>and</strong> nutrients<br />
were added to the aerated chamber <strong>and</strong> the pH was adjusted. Because the concentration of<br />
suspended microbes was exceptionally high <strong>and</strong> because no washout of activated sludge<br />
could occur, BOD removal exceeded 99%.<br />
In some other aerobic treatment plants the biomass is attached to a supporting solid,<br />
<strong>and</strong> is in contact with the aerated effluent. The oldest of this type of system is the<br />
trickling bed filter, in which effluent is sprayed over the surface of a bed of rough solids<br />
(such as gravel, broken rocks, or coke) <strong>and</strong> trickles downward over the solid's surfaces,<br />
meeting an up-flow of air. The beds may be circular or rectangular in plan <strong>and</strong> 2 3 m<br />
(6.56 9.84 ft.) deep. The bed packing becomes coated with a very complex mixture of<br />
microbes <strong>and</strong> other organisms, which oxidize dissolved organic materials <strong>and</strong> reduce the<br />
BOD. As one pass is insufficient the liquid is recirculated. At intervals surplus sludge<br />
sloughs away <strong>and</strong> is collected in a settling tank. To cope with BOD-rich effluents several<br />
filters may operate in sequence. Such filters are expensive to build, they occupy a large<br />
area <strong>and</strong> they have a limited capacity, being liable to `ponding' if overloaded. Sometimes<br />
problems arise from flies which multiply on the biomass. These filters have increasingly<br />
been replaced by high-rate biofiltration towers. These towers, which may be 4.3 6.1 m<br />
(14 20ft.) high, are packed with plastic units with shapes designed to support the<br />
biomass <strong>and</strong> to have a large surface to volume ratio but to be resistant to blockage.<br />
Effluent is sprayed into the top of the tower <strong>and</strong> as it trickles downward it meets an upflow<br />
of air <strong>and</strong> is oxidized by the film of microbes on the plastic. Towers have capacities<br />
about ten times those of trickling filters covering the same areas. If desired the liquid can<br />
be re-circulated <strong>and</strong> a BOD removal of 60 65% is obtained, so they achieve a partial<br />
treatment. Towers are followed by settling tanks, which retain sludge. If followed by<br />
further treatment in a low-loaded activated sludge plant 97% BOD removal <strong>and</strong> 94% SS<br />
removal has been achieved. If towers are overloaded they, like trickling filters, can give<br />
rise to unpleasant odours.<br />
Rotating disc contactors are another type of plant in which most of the biomass is<br />
supported on sets of lightweight plastic discs mounted side by side along a rotating axle.<br />
The surfaces of the slowly rotating discs are alternately immersed in a trough of effluent<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>and</strong> then, together with the film of liquid coating the surface, are exposed to air. Strips<br />
projecting from the faces of the discs form chambers, which cause air to be carried down<br />
below the surface of the liquid (KuÈhtreiber <strong>and</strong> Laa-Thaya, 1995). Discs are about 45%<br />
submerged. Much of the biomass is supported on the discs but some may float free in the<br />
liquid in the trough. Other forms of rotating contactors have also been used. The energy<br />
requirements of these devices are much lower than those using compressed air for<br />
aeration. BOD removal of 80 90% is possible.<br />
Other types of aerobic treatment plants have been tried. For example, fluidised beds<br />
have been tested <strong>and</strong> some are in use (Section 3.2.4, Fig. 3.1). In these the biomass is<br />
supported in agranular, porous <strong>and</strong> relatively dense material. The mass, lifted into the<br />
flowing effluent by the aeration bubbles, achieves ahigh biomass density because the<br />
material supporting the biomass is too dense to be swept out of the aeration chamber,<br />
removingthe needforasettlementtank.Inanotherapproach,inapilotplant,the biomass<br />
was supported in small pieces of plastic sponge (Leeder, 1986). When these were<br />
overloaded with biomass the surplus was removed by collecting the pieces of sponge <strong>and</strong><br />
squeezing them between rollers.<br />
3.8.3 Sludge treatments <strong>and</strong> disposal<br />
Aerobic effluent treatments inevitably produce considerable amounts of surplus biomass,<br />
`sludge', which retains about 70% of the mass of the substances that contribute to the<br />
BOD. The treatment <strong>and</strong> disposal of this sludge is inconvenient <strong>and</strong> contributes<br />
substantially, about 50%, to the cost of the treatment (Armitt, 1981; Huige, 1994; Vriens<br />
et al., 1986, 1990). Sludges differ in their characteristics, such as the ease with which<br />
they will settle <strong>and</strong> how compact they are. Normally sludges are collected by<br />
sedimentation when, with the addition of coagulants, they may have asolids content<br />
of 1 2% dry matter. After afurther period of settling the solids content will increase to,<br />
say, 2 4%. After each concentration treatment the liquid that has separated is returned to<br />
the effluent treatment plant. The sludge may be consolidated further by centrifugation,<br />
vacuum filtration or in a pressure filter. Each consolidation reduces the volume of sludge<br />
to be h<strong>and</strong>led <strong>and</strong> so reduces the transport costs as the material is carted away for disposal<br />
(Table 3.5). An alternative is to treat the sludge with coagulants followed by flotation <strong>and</strong><br />
collection by skimming. In some circumstances the sludge may be `stabilized', for<br />
example by aeration or by anaerobic digestion. Both of these treatments reduce the bulk.<br />
Anaerobic digestion is accompanied by the generation of methane-rich biogas, which can<br />
be used as a fuel. It is doubtful if this technique is used by brewers. It is used at some<br />
large sewage works. In some places dried sludge is incinerated, but again this does not<br />
seem to be suitable for brewers unless they are in a hot <strong>and</strong> dry region where the material<br />
can be dried spread on earth beds to dry in the open. Brewers usually have the sludge<br />
Table 3.5 The influence of dewatering on the volume of sludge (Lloyd, 1981)<br />
Stage of dewatering Moisture content (%) Equivalent volume<br />
From settling tanks 98.5 100<br />
After further settling 97.0 50<br />
After decantation 95.5 33<br />
After centrifugation 80.0 7<br />
From vacuum filter 70.0 5<br />
From filter press 65.0 4<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
emoved by contractors who will bury it in l<strong>and</strong>fill sites or may arrange to have it spread<br />
on farml<strong>and</strong>. The latter method of disposal is only available at certain times of the year.<br />
Because of its composition the sludge is potentially valuable. Markets for it have been<br />
sought, <strong>and</strong> it has been used (with or without admixture with lime) as a soil conditioner<br />
<strong>and</strong> has been used as a supplement in animal feeds.<br />
3.8.4 Anaerobic <strong>and</strong> mixed treatments of brewery effluents<br />
It was once considered impracticable to treat brewery effluents by anaerobic digestion,<br />
but methods for doing this have been developed (Anderson <strong>and</strong> Saw, 1986; Benson et al.,<br />
1997; Driessen et al., 1997; Eder, 1982; Fuchs, 1995; Gerards <strong>and</strong> Vriens, 1996;<br />
Hellriegel, 1996; Huige, 1994; Klijnhout <strong>and</strong> Van Eerde, 1986; Langereis <strong>and</strong> Smith,<br />
1998; Love, 1987; Martin <strong>and</strong> Sanchez, 1987; Mayer <strong>and</strong> Eeckhaut, 1997; Pipyn et al.,<br />
1983; Schumann, 1999; Schur et al., 1995; Swinkels et al., 1985; Vriens et al.,1986,<br />
1990). The effluent supplied to an anaerobic digestion plant must be carefully regulated<br />
in terms of its pH, flow, temperature, <strong>and</strong> BOD. These plants operate best on a steady<br />
flow of BOD-rich effluents <strong>and</strong> they are easily put out of commission by `shocks' or<br />
traces of toxic substances, so preliminary mixing <strong>and</strong> buffering treatments must have<br />
been applied to the effluent input. Start-up times are slow (4 10 weeks) because the<br />
anaerobic micro-organisms multiply slowly. Despite these stringent requirements<br />
anaerobic plants are being used both because they produce very little sludge, <strong>and</strong><br />
because they are comparatively compact (about 70% less space than aerobic plant) <strong>and</strong><br />
cheap to construct <strong>and</strong> run (no aeration plant is needed) <strong>and</strong> because biogas is generated<br />
(0.3 0.5 m 3 /kg COD removed) <strong>and</strong> this may be used as fuel in the boiler house.<br />
Anaerobic systems never completely remove BOD (50 95%, usually 70 85%), <strong>and</strong><br />
COD removal is generally 60 75%. Essentially no nitrogen or phosphate is removed by<br />
anaerobic treatments. They are best regarded as preliminary treatments for strong<br />
effluents that must be followed by an aerobic treatment for weak effluents, separated<br />
from the strong effluents at the brewery, <strong>and</strong> the effluent from the anaerobic treatment.<br />
This may occur either at the brewery or at the sewage works.<br />
Anaerobic treatments occur in three stages, but often the first two take place in a<br />
single vessel. In the first stage, which may be aerobic, microbes generate <strong>and</strong> release<br />
hydrolytic enzymes that degrade the complex molecules present in the effluent to<br />
smaller molecules, e.g., polysaccharides to simple sugars, proteins to amino acids <strong>and</strong><br />
free fatty acids are liberated from lipids. These simple molecules are easily assimilated<br />
by microbes. In the second, or `acidification' stage, which is strictly anaerobic, many of<br />
these molecules are converted to organic acids. At this stage some biogas, containing<br />
hydrogen <strong>and</strong> carbon dioxide, is produced. The products of the acidification stage are<br />
good substrates for the microbes needed in the next stage, but the best cultural conditions<br />
are different. After the acidification process the pH (6.6 7.6), temperature, <strong>and</strong> nitrogen<br />
<strong>and</strong> phosphate levels are adjusted as the liquid flows to the `methanization' vessel.<br />
Methanization is a strictly anaerobic process. It may be carried out at ambient<br />
temperatures (18 20 ëC; 64.4 68 ëF) or with thermophilic organisms at about 50 ëC<br />
(122 ëF) but in <strong>practice</strong> mesophilic conditions are usually chosen (about 35 ëC, 95 ëF).<br />
The effluent to be treated may be warmed by steam, partly produced by burning the<br />
biogas, <strong>and</strong> partly by heat recovered from the brewery. The biogas produced contains<br />
55 75% methane, 1 5% hydrogen, 25 40% carbon dioxide <strong>and</strong> 1 7% nitrogen, as<br />
well as traces of ammonia <strong>and</strong> hydrogen sulphide. This gas needs to be scrubbed with<br />
sodium hydroxide to remove the carbon dioxide <strong>and</strong> this may need to be supplemented<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
with hydrogen peroxide to oxidize the hydrogen sulphide. Alternatively, the sulphide<br />
may be removed as iron sulphide or be adsorbed by silica gel. Escaping gases may be<br />
deodorized by passage through abiofilter.<br />
Variouskindsofequipmentareinusefortheanaerobic digestionofbrewery effluents.<br />
Stirred vessels have been used, with or without the microbes being supported on solid<br />
carrier materials. The preferred systems now use the UASB or upward-flow anaerobic<br />
sludge blanket system in which acidified effluent iscontinuously mixed into the base of a<br />
reaction vessel. As the liquid rises it passes through alayer of agglomerated microbes<br />
which have formed spheres 2 5mm in diameter. The process automatically selects for<br />
microbeswhichwillclumptogether.Theorganicmaterialsinsolutionareattackedbythe<br />
microbes, which release biogas. Very little extra sludge is formed. The biogas rises,<br />
carrying some of the granular microbial blanket. Towards the top of the vessel the rising<br />
mixture meets athree-phase separator. The gas is collected <strong>and</strong> taken from the vessel <strong>and</strong><br />
the granular microbial material, which had been carried up by the gas bubbles, settles<br />
back into the body of the reactor. The liquid flows up <strong>and</strong> over one or more weirs <strong>and</strong><br />
leaves the vessel. After passing through ade-foamer <strong>and</strong> being scrubbed the biogas, now<br />
chiefly methane, may be stored in agas-holder, be flared off or be directed to aboiler. In<br />
one case biogas provided 8% of the boiler's fuel requirements. The reactor requires a<br />
constant flow of liquor so if there is acheck in the inflow of effluent the liquor is recirculated.<br />
It is advantageous to have two reactors working in series.<br />
Anexampleofananaerobic<strong>and</strong>aerobicplantforpartialeffluenttreatmentisshownin<br />
Fig. 3.2. This plant, which was designed to occupy asmall ground area, has anumber of<br />
novel features (Driessen et al., 1997; Meijer, 1998). The IC Õ reactor is atwo-stage<br />
methanization tower, which can be regarded as having two UASB reactors mounted one<br />
above the other. In the aerobic, Circox Õ tower the microbes are supported on grains of<br />
Fig.3.2(opposite) Diagramofananaerobic/aerobiceffluenttreatmentplant(afterDriessenetal.,<br />
1997). The fully enclosed plant is situated on a confined site (200 m 2 ) in a built-up area. Screened<br />
waste water (design flow of up to 4,200 m 3 /day) is directed to a buffer tank (500 m 3 ; 25 m high) or,<br />
in the case of alkaline or other `extreme' wastes, to a calamity tank (150 m 3 ), from where it may be<br />
metered into the buffer tank or to the sewer. Mixed effluent is transferred to the pre-acidification<br />
(PA) tank (500 m 3 ; 25 m high) where acidification <strong>and</strong> the hydrolysis of polymeric materials occur<br />
<strong>and</strong> where the pH <strong>and</strong> nutrient levels are adjusted. Phosphoric acid <strong>and</strong> urea may be added if<br />
required. The acidified effluent goes to the base of the IC Õ reactor tower (390 m 3 ; 20 m high). This<br />
is, in effect, two strictly anaerobic UASB reactors mounted one above the other. The influent enters<br />
the base (6) of the high-loaded reactor (1) <strong>and</strong> is mixed in with the surrounding fluid <strong>and</strong> granular<br />
microbial sludge. It rises through the sludge blanket in the reactor <strong>and</strong> meets the first three-phase<br />
separator (3). The gas rises to a de-foaming unit generating a gas-lift, <strong>and</strong> the liquid <strong>and</strong> entrained<br />
sludge return to (6) via a `downer' pipe, (5). The sludge, from the first three-phase separator, settles<br />
back into the body of the reactor while the liquid rises into the second, low-loaded reactor (2),<br />
moves through the second sludge blanket <strong>and</strong> meets the second three-phase separator (4). The<br />
biogas is scrubbed, collected <strong>and</strong> utilized, the sludge is retained in the reactor <strong>and</strong> the effluent<br />
moves on to the Circox Õ reactor (230 m 3 ; 19 m high). This second reactor is an aerobic tower, with<br />
an internal circulation through the inner, `riser' pipe (7) <strong>and</strong> the outer, annular `downer' channel.<br />
The circulation is maintained by the air-lift generated by the aeration air injected into the base. The<br />
ventilation air is drawn from all parts of the plant <strong>and</strong> sulphides in the air are oxidized to sulphate.<br />
The biomass is supported on granular basalt, so it readily sediments when it reaches the settling<br />
space (10). The gases rise through a pipe (9), <strong>and</strong> are vented. The effluent rises over weirs, which<br />
retain the biomass, <strong>and</strong> is directed by way of a cyclone tank that retains some suspended solids, but<br />
not fine suspended matter, like kieselguhr, which passes with the liquid into the sewer. The system<br />
is compact <strong>and</strong> has a rapid throughput with a low production of surplus sludge. The average<br />
reductions of the total COD <strong>and</strong> the soluble COD are 80% <strong>and</strong> 94% respectively.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Waste water<br />
Screen<br />
0.5 mm<br />
Ventilation air<br />
Solid<br />
waste<br />
NaOH<br />
HCl<br />
500 m 3<br />
Buffer<br />
tank<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Calamity<br />
tank<br />
To<br />
sewer<br />
500 m 3<br />
Preacidification<br />
tank<br />
Biogas<br />
4<br />
2<br />
3<br />
1<br />
Defoam<br />
tank<br />
IC reactor<br />
390 m 3<br />
Air<br />
5<br />
Gas/liquid<br />
separator<br />
Effluent<br />
6<br />
Gas<br />
scrubber<br />
10<br />
7<br />
Gasholder<br />
10 m 3<br />
Vented air<br />
9<br />
10<br />
8<br />
Aerobic<br />
Circox<br />
230 m 3<br />
Flare<br />
320 m 3<br />
Cyclone<br />
tank<br />
Effluent<br />
to sewer<br />
Biofilter<br />
14 m 2<br />
Effluent<br />
to sewer
asalt, which readily settle <strong>and</strong> so preventing the biomass being washed out of the<br />
column. The plant removes 80% of the total COD <strong>and</strong> 94% of the soluble COD. In this<br />
particular plant the kieselguhr is not removed <strong>and</strong> is discharged in the final effluent,<br />
which is delivered to the sewer. Other combined anaerobic <strong>and</strong> aerobic plants have been<br />
described (e.g. Eyben et al., 1985, 1995; Gerards <strong>and</strong> Vriens, 1996; KuÈhbeck, 1995;<br />
Mayer <strong>and</strong> Eeckhaut, 1997; Vriens et al., 1990). Under aerobic conditions microbes take<br />
up phosphate, <strong>and</strong> many store it as polyphosphate, <strong>and</strong> so this is removed from the<br />
effluent. The situation with nitrogen removal is more complex (Vriens et al., 1990).<br />
Aerobic conditions are needed to oxidize ammonium ions to nitrate, then anaerobic<br />
conditions are required for the nitrate to be reduced to nitrogen gas which is removed<br />
from the system (Eyben et al., 1995).<br />
3.9 Other water treatments<br />
Reedbeds have been used to treat some crude or partially purified malting effluents or<br />
even some sludge (Maule et al., 1996; Walton, 1995). They are essentially shallow tanks<br />
filled with a permeable support such as limestone chips on which reeds, e.g., Phragmites<br />
spp. <strong>and</strong> perhaps yellow flag iris are planted. The effluent percolates through the beds,<br />
among the roots of the plants. Suspended solids have been reduced from 80 to less than<br />
20 mg/l, <strong>and</strong> COD from 60 to less than 10 mg/l. The removal of nitrogen <strong>and</strong> phosphate<br />
appears to be satisfactory. A problem with reed beds is that they occupy large areas of<br />
ground.<br />
Treated effluents have been polished in various ways, including passage through reed<br />
beds, passage through large, shallow lagoons (with or without forced aeration), <strong>and</strong><br />
recycling through percolating filters. As the cost of acquiring water continues to rise <strong>and</strong><br />
the pressure to treat effluents to higher st<strong>and</strong>ards continues to increase it is inevitable that<br />
brewers working in large breweries will frequently consider purifying their effluents<br />
sufficiently to allow the water to be recovered <strong>and</strong> used, at least for some purposes, so<br />
minimizing water acquisition <strong>and</strong> effluent disposal charges. The re-use of effluent is<br />
likely to involve a two-stage aerobic or an anaerobic/aerobic treatment, followed by a<br />
polishing treatment, sterilization, s<strong>and</strong> filtration possibly followed by finer filtration,<br />
carbon filtration <strong>and</strong> perhaps ultrafiltration or demineralization by ion exchange.<br />
3.10 References<br />
ANDERSON, G. K. <strong>and</strong> SAW, C. B. (1986) Pauls <strong>Brewing</strong> Room Book, 1986±1988 (8th edn), Pauls' Malt,<br />
Ipswich, p. 37.<br />
ANDERSSON, L. E. <strong>and</strong> NORMAN, H. (1997). Brew. Distill. Internat., 28 (7), 18.<br />
ANON. (1988) European Brewery Convention Monograph XIV. EBC Symposium. Water in the <strong>Brewing</strong><br />
Industry, Zoeterwoude.<br />
ARMITT, J. D. G. (1981) in <strong>Brewing</strong> Science 2 (Pollock, J. R. A., ed.) Academic Press, London, p. 551.<br />
ASKEW, M. (1975) Process Biochem., 10 (1), 5.<br />
ASKEW, M. (1987), Brewer, 73, Nov., p. 500.<br />
ASKEW, M. <strong>and</strong> ROGERS, S. (1997), Brewers' Guard., 126 (5), 24.<br />
BAK, S. N., EKENGREN, OÈ ., EKSTAM, K., HAÈ RNULV, G., PAJUNEN, E., PRUCHA, P. <strong>and</strong> RASI, J. (2001) European<br />
Brewery Convention Manual of Good Practice. Water in <strong>Brewing</strong>. Hans Carl, NuÈrnberg, 128 pp.<br />
BAXTER, E. D. <strong>and</strong> HUGHES, P.S. (2001) Beer: quality, safety <strong>and</strong> nutritional aspects. Cambridge. The<br />
Royal Society of Chemistry, 138 pp.<br />
BENSON, J. T., COLEMAN, A. R., DUE, J. E. B., HENHAM, A. W., TWAALFHOVEN, J. G. P. <strong>and</strong> VINCKX, W. (1997)<br />
European Brewery Convention Manual of Good Practice: Brewery Utilities. Hans Carl, NuÈrnberg,<br />
p. 137.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
BERKMORTEL, H. A. VAN DEN. (1988a) European Brewery Convention Monograph XIV. EBC Symposium.<br />
Water in the <strong>Brewing</strong> Industry, Zoeterwoude, p. 49.<br />
BERKMORTEL, H. A. VAN DEN. (1988b) MBAA Tech Quart., 25 (3), 85.<br />
BLACKMANN, B. (1998) Brewers' Guard. 127, Apr., p. 30.<br />
BRAUN, G. <strong>and</strong> NIEFIND, H.-J. (1988) European Brewery Convention Monograph XIV. EBC Symposium.<br />
Water in the <strong>Brewing</strong> Industry, Zoeterwoude, p. 68.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting <strong>and</strong> <strong>Brewing</strong> Science, Vol I,<br />
Malt <strong>and</strong> Sweet Wort (2nd edn). Chapman <strong>and</strong> Hall, London. p. 194.<br />
BROOKS, R. B., HALFORD, M. H. <strong>and</strong> SKINNER, R. N. (1972) Proc. 12th Conv., Inst. of <strong>Brewing</strong> (Australia<br />
<strong>and</strong> New Zeal<strong>and</strong> Sect.), Perth, p. 73.<br />
BROWN, J. W., BOTT, N.J., BOWERMAN, R. <strong>and</strong> SMITH, D. (1999) Proc. 27th Congr. Eur. Brew. Conv.,<br />
Cannes, p. 225.<br />
BYRNE, H. (1990) Ferment, 3 (2), 90.<br />
CLEATHER, T. G. (1992) Proc. 5th Internat. Brew. Tech. Conf., Harrogate, p. 287.<br />
COMRIE, A. A. D. (1967) J. Inst. <strong>Brewing</strong>, 73, 335.<br />
COORS, J. H. <strong>and</strong> JANGAARD, N. O. (1975) Proc. 16th Congr. Eur. Brew. Conv., Nice, p. 311.<br />
COWAN, J. A. <strong>and</strong> WESTHUYSEN, J. VAN DER (1999) Proc. 7th Sci. Tech. Conv. Inst. <strong>Brewing</strong>, Africa Sect.,<br />
Nairobi, p. 193.<br />
CRISPIN, P. (1996) Brewers' Guard., 125 (9), 34.<br />
DRIESSEN, W., HABETS, L. <strong>and</strong> VEREIJKEN, T. (1997) Ferment, 10 (4), 243.<br />
EDEN, G. E. (1987) Brewer, 73, Nov., 504.<br />
EDER, L. J. (1982) MBAA Tech Quart. 19 (3), 111, 138.<br />
EUMANN, M. (1999) Proc. 7th Sci. Tech. Conv. Inst. <strong>Brewing</strong>, Africa Sect., Nairobi, p. 169.<br />
EYBEN, D., GERARDS, R. <strong>and</strong> VRIENS, L. (1995) MBAA Tech. Quart., 32 (3), 142.<br />
EYBEN, D., VRIENS, L., FRANCO, P. <strong>and</strong> VERACHTERT, H. (1985) Proc. 20th Congr. Eur. Brew. Conv.,<br />
Helsinki, p. 555.<br />
FUCHS, C. B. (1995) MBAA Tech. Quart., 32, 85.<br />
GERARDS, R. <strong>and</strong> VRIENS, L. (1996) Proc. 24th Conv. Inst. <strong>Brewing</strong>, Asia-Pacific Section, Singapore,<br />
p. 192.<br />
GOUGH, A. J. E. (1995) MBAA Tech. Quart., 32, 195.<br />
GRANT, A. P. (1995) Ferment, 8 (4), 252.<br />
HACKSTAFF, B. W. (1978) MBAA Tech. Quart., 15 (1), 1.<br />
HARRISON, J. G., LAUFER, S., STEWART, E. D., SIEBENBERG, J. <strong>and</strong> BRENNAN, M. W. (1963) J. Inst. <strong>Brewing</strong>,<br />
69 (4), 323.<br />
HARTEMANN, P. (1988) European Brewery Monograph XIV. EBC Symposium, Water in the <strong>Brewing</strong><br />
Industry, Zoetewoude. p. 37.<br />
HELLRIEGEL, K. (1996) Brauwelt Internat., 14 (5), 422.<br />
HERON, J. R. (1989) Ferment, 2 (2), 118.<br />
HORRIGAN, R., LLOYD, W. J. W. <strong>and</strong> YOUNG, I. M. (1989) Project No. 59. Report of the Joint Maker/User<br />
Committee, Inst. <strong>Brewing</strong> <strong>and</strong> ABTA. Conservation of water <strong>and</strong> reduction of effluent.<br />
HUGHES, D. A. (1987) Brewer, 73 (872), 266.<br />
HUIGE, N. J. (1994) in H<strong>and</strong>book of <strong>Brewing</strong> (Hardwick, W. A., ed.), Marcel Dekker, New York. p. 501.<br />
IBBOTSON, G. E. (1986) Brewer, 72 (859), 169.<br />
ISAAC, P. G. (1976) Process Biochem. 11 (2), 17.<br />
ISAAC, P. C. G. <strong>and</strong> ANDERSON, G. K. (1973), J. Inst. <strong>Brewing</strong>, 79, 154.<br />
KLIJNHOUT, A. F. <strong>and</strong> VAN EERDE, P. (1986) J. Inst. <strong>Brewing</strong>, 92, 426.<br />
KUÈ HBECK, G. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, p. 751.<br />
KUÈ HTREIBER, F. <strong>and</strong> LAA-THAYA, A. (1995) Brauwelt Internat. (1), 45.<br />
KUNZE, W. (1996) Technology <strong>Brewing</strong> <strong>and</strong> Malting, VLB, Berlin, p. 60.<br />
LANGEREIS, W. H. <strong>and</strong> SMITH, C. G. (1998) Proc. 25th Conv. Inst <strong>Brewing</strong>, Asia-Pacific Sect., Perth,<br />
p. 187.<br />
LEEDER, G. I. (1986) Brewer, 72, p. 213.<br />
LLOYD, W. J. W. (1981) Brewer, 67, 396.<br />
LOM, T. <strong>and</strong> FEDDERSON, C. C. (1981) Proc. 18th Congr. Eur. Brew. Conv., Copenhagen, p. 121.<br />
LOVE, L. S. (1987) MBAA Tech. Quart., 24 (2), 51.<br />
LUNNEY, M. J. (1981) Brewers' Guard, 110 (1), 13.<br />
MAILER, A., PEEL, R. G., THEAKER, P. D. <strong>and</strong> RAVINDRAN, R. (1989) MBAA Tech. Quart., 26, 35.<br />
MARTIN, S. <strong>and</strong> SANCHEZ, R. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid, p. 655.<br />
MAULE, A. P., TRUMPESS, C. R. <strong>and</strong> THURGOOD, S. D. (1996) Proc. 6th Internat. Brew. Tech. Conf.,<br />
Harrogate. <strong>Brewing</strong> Technology, the Market <strong>and</strong> the Environment. p. 408.<br />
MAYER, W. <strong>and</strong> EECKHAUT, M. (1997) Brauwelt Internat., 15 (5), 414.<br />
MCGARRITY, M. J. (1990) Louvain <strong>Brewing</strong> Lett., 4 (3/4), 3.<br />
MEIJER, D. (1998) Brauwelt Internat., 16, 455.<br />
MEYER, H. (1973) Proc. 14th Congr. Eur. Brew. Conv., Salzburg, p. 429.<br />
MOLL, M. (1979) in <strong>Brewing</strong> Science, 1 ( Pollock, J. R. A., ed.) London, Academic Press. p. 539.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
MOLL, M. (1995) in A H<strong>and</strong>book of <strong>Brewing</strong> (Hardwick, W. A. ed.). New York, Marcel Dekker, p. 133.<br />
PENROSE, J. D. F. (1985) Brewers' Guard., 114, 25.<br />
PIPYN, P., OMBREGT, J. P. <strong>and</strong> TOYE, J. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 587.<br />
PUTMAN, R. (2001) Brewer Internat., 2 (4), 27.<br />
REED, R. <strong>and</strong> HENDERSON, G. (1999/2000) Ferment 12 (6), 13.<br />
ROBERTSON, J. L., BROWN, L. C. <strong>and</strong> MURPHY, K. L. (1979) MBAA Tech. Quart. 16 (1), 33.<br />
ROSTRON, J. (1996) Brewers' Guard., 125, (5), 20.<br />
RUDIN, A. D. (1976) Brewers' Guard., 105 (12), 30.<br />
SCHUMANN, G. (1999) Brauwelt Internat., 17 (5), 374.<br />
SCHUR, F., BHEND, D., BUCHER, A. J. <strong>and</strong> WETZEL, E. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels,<br />
p. 741.<br />
SWINKELS, K. T. M., VEREIJKEN, T. L. F. <strong>and</strong> HACK, P. J. F. (1985) Proc. 20th Congr. Eur. Brew. Conv.,<br />
Helsinki, p. 563.<br />
TAYLOR, D. (1981) Brew. Distill. Internat., 11, 35, 42.<br />
TAYLOR, D. G. (1989) Ferment, 2 (1), 76.<br />
THEAKER, P. D. (1988) Brewers' Guard., 117 (3), 22.<br />
THOMPSON, J. (1995) Ferment, 8 (3), 177.<br />
VEREIJKEN, T. L. F., DRIESSEN, W. J. B. <strong>and</strong> YSPEART, Y. (1999) Proc. 7th Sci. Tech. Conf. Inst. <strong>Brewing</strong>,<br />
Africa Sect., Nairobi, p. 174.<br />
VRIENS, L., EYBEN, D. <strong>and</strong> VERACHTERT, H. (1986) Proc. Symp. J. De Clerck Chair II; Microbiology <strong>and</strong><br />
the <strong>Brewing</strong> Industry from Barley to Beer. Leuven/Louvain, p. 128.<br />
VRIENS, L., VAN SOEST, H. <strong>and</strong> VERACHTERT, H. (1990) CRC Crit. Revs. Biotechnol., 10 (1), 1.<br />
WALKER, M. J. (1994) Proc. 4th Aviemore Conf. On Malting, <strong>Brewing</strong> <strong>and</strong> Distilling. London, Inst of<br />
<strong>Brewing</strong>, p. 345.<br />
WALTON, C. (1995) Brew. Distill. Internat., 26 (7), 29.<br />
WARD, J. A. (2000) Proc. 26th Conv. Inst. <strong>Brewing</strong>, Asia-Pacific Sect., Singapore, p. 122.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
4<br />
The <strong>science</strong> of mashing<br />
4.1 Introduction<br />
Mashing consists of mixing ground malt (usually amixture of malts) <strong>and</strong> other prepared<br />
grist materials (appropriate adjuncts <strong>and</strong> sometimes salts <strong>and</strong> sometimes supplementary<br />
enzymes,Chapter2)withacarefullycontrolledamountofliquoratachosentemperature.<br />
In some few instances the mash may be made with mainly, or entirely, unmalted<br />
preparations of cereals mixed with industrial enzymes. The liquor is nearly always prepurified<br />
<strong>and</strong> contains achosen mixture of dissolved salts (Chapter 3). In the simplest<br />
systems, after a `st<strong>and</strong>' or st<strong>and</strong>s of various durations, at one or more selected<br />
temperatures, the liquid, or`sweet wort' isseparatedfrom the residual solids of the mash,<br />
the spent grains or draff. In other cases portions of the mash are withdrawn <strong>and</strong> heated<br />
before being added back to the `main mash' while in other cases cereal preparations,<br />
cooked in a separate vessel, are transferred <strong>and</strong> mixed into the main malt mash.<br />
Representative mashing schedules are considered below. The sweet wort goes forward to<br />
the kettle or copper, to be boiled with hops, while the spent grains are disposed of<br />
(Chapter 3). The, apparently simple, processes of mashing conceal a very complex<br />
mixture of physical, chemical <strong>and</strong> biochemical changes. An underst<strong>and</strong>ing of these<br />
changes has been essential to permit the logical development of mashing conditions for<br />
the preparation of desirable <strong>and</strong> uniform worts in rapid <strong>and</strong> reproducible ways. Thus the<br />
purpose of mashing is economically to prepare wort of the correct composition, flavour<br />
<strong>and</strong> colour in the highest practical yield <strong>and</strong> in the shortest time.<br />
Thewortis partly characterized by its`strength', the amount ofsolids, or`extract'that<br />
are in solution <strong>and</strong> the volume of liquid in which the solids are dispersed. Unfortunately,<br />
the concentration of wort is expressed in avariety of different units (Appendix). In the<br />
simplest case the specific gravity is used as a measure. The higher the specific gravity the<br />
more concentrated the solution of wort solids. For example, a wort might have a specific<br />
gravity (s.g.) of 1.040 at 20 ëC (68 ëF), relative to pure water at 1.000, or the same wort<br />
has an SG of 1040 relative to water as 1000.0. In other systems the amounts of solids in<br />
solution are estimated from the s.g. <strong>and</strong> reference to tables relating to the s.g. values of<br />
sucrose solutions, either from the table of Balling (ASBC) or the table of Plato (EBC; see<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Appendix). So awort with an SG of 1040 has aconcentration of solids of about 9.99%<br />
(w/w) assuming that the wort solids influence the SG in an identical way to sucrose. In<br />
the old British system (pre-1977) the extract in a barrel of wort at 15.5 ëC (60 ëF) was the<br />
excess weight of a barrel of wort in pounds (lb.) over the weight of a barrel of water,<br />
360 lb. So the SG ˆ (excess Brewer's lb. + 360)/360. So in the case given, SG ˆ 1040 (at<br />
15.5 ëC), the excess SG ˆ 40 so the extract, in Brewer's lb. ˆ 40 0.36 ˆ 14.4. The<br />
efficiency of mashing is often estimated by comparing the extract recovered in the<br />
brewery with that obtained in laboratory mashes when the hot water extract (HWE) of the<br />
grist is determined. Reliable estimates of extract recoveries are not freely available, but<br />
while in old mash tuns the value might be about 85 95%, in newer mash filters the value<br />
may equal or just exceed 100%, a commercially valuable advantage provided that wort<br />
quality is maintained.<br />
The importance of enzymes in mashing is illustrated by the fact that a cold water<br />
extract of a pale malt (CWE; preformed soluble materials), prepared in a cool, dilute<br />
solution of ammonia to stop enzyme activity, will be in the range 15 22%, dry basis,<br />
while the hot water extract (HWE) will be 75 83%, dry basis. Thus, by holding a smallscale<br />
mash for between one <strong>and</strong> two hours under conditions such that enzyme activity is<br />
favoured, some 53 68% more of the malt solids are brought into solution as the result of<br />
enzyme-catalysed reactions. While about half of the CWE solids are fermentable by<br />
yeasts typically 75% <strong>and</strong> up to about 85% of the HWE is fermentable. The enzyme<br />
catalysed changes that occur during mashing are more complex than those normally<br />
investigated by biochemists, who usually study each enzyme acting in isolation, with a<br />
homogeneous substrate, at one temperature, with an unchanging pH <strong>and</strong> with a large<br />
excess of substrate to maintain enzyme activity.<br />
In contrast, in mashing, a very large number of enzymes act simultaneously on the<br />
components of the grist (malt <strong>and</strong> mash tun adjuncts) under conditions that are far from<br />
optimal for many of them in terms of substrate concentration <strong>and</strong> accessibility, pH <strong>and</strong><br />
enzyme stability. Enzymes are inactivated at different rates depending on the<br />
temperature, the pH, the presence of substrate <strong>and</strong> other substances (such as tannins<br />
<strong>and</strong> cofactors such as calcium ions) in solution. Starch, proteins, nucleic acids, lipids <strong>and</strong><br />
other substances are attacked, usually by hydrolytic reactions, but other reactions, such as<br />
oxidations, also occur. Not only are enzymes progressively inactivated but substrate<br />
concentrations alter <strong>and</strong>, in the case of starch for instance, are nearly totally degraded.<br />
Solid starch granules are not readily degraded until gelatinization temperatures are<br />
approached <strong>and</strong> starch grains are disrupted. The polysaccharide cell walls <strong>and</strong> the starch<br />
granules are coated with proteins that seem to impede their enzymic degradation. Where<br />
grist particles are relatively large <strong>and</strong> the cell walls are intact, as in unmodified fragments<br />
of malt <strong>and</strong> some adjuncts, the cell walls prevent enzymes reaching <strong>and</strong> degrading the<br />
cell's contents. In many instances the products of hydrolysis competitively inhibit<br />
enzyme activity <strong>and</strong> in a number of instances (e.g. -amylase, limit dextrinase <strong>and</strong> some<br />
proteolytic enzymes) proteins occur which partially or largely inhibit enzyme activities.<br />
In addition some enzymes occur bound to insoluble materials in the mash, preventing<br />
them diffusing <strong>and</strong> so limiting their ability to reach their substrates. This is true of -<br />
amylase, proteases <strong>and</strong> -glucosidase in malts. The significance of insoluble enzymes<br />
<strong>and</strong> the presence of endogenous enzyme inhibitors in mashes has been widely ignored<br />
A knowledge of the properties of enzymes is essential for an underst<strong>and</strong>ing of mashing<br />
regimes, yet the traditional methods for studying their properties are of limited use.<br />
Brewers make use of the concept of `optima' in considering enzyme activities in mashes.<br />
This is helpful, but it must be realized that a temperature or a pH optimum is not a true<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
constant. Under any particular set of conditions an enzyme may be stable <strong>and</strong> so, when it<br />
is incubated with asubstrate, product(s) will be formed at aconstant rate until substrate<br />
concentration is reduced to asignificant extent. If the temperature is increased then the<br />
reaction catalysed will proceed faster. However, if at this higher temperature the enzyme<br />
is progressively inactivated then the rate of formation of the product(s) will decline <strong>and</strong><br />
will ultimately stop even if some substrate remains. At progressively higher temperatures<br />
the reaction rates catalysed by the `native', undenatured enzyme will increase, but the<br />
rate of enzyme inactivation also increases <strong>and</strong> so the production of product will cease<br />
sooner. Thus the amount of product formed, at aparticular temperature in agiven time,<br />
will depend on the rates of catalysis <strong>and</strong> on the rate of enzyme inactivation. The longer<br />
the reaction period the lower the optimum temperature, that is, the temperature at which<br />
the most product will have been produced (Fig. 4.1). In mashing, the rate of appearance<br />
of asubstance depends on the mixture in the grist, the mash thickness, the fineness of the<br />
grind <strong>and</strong> so the particle size distribution in the grist. Similarly the optimum pH of a<br />
reaction varies with the test conditions. The situation with pH is complicated by the<br />
frequentfailure totakeaccountofthepHchangesthatoccurasthetemperatureisaltered,<br />
<strong>and</strong> the difficulty of measuring pH at the elevated temperatures used in mashing (see<br />
below).<br />
Sweet wort is viscous, sweet, dense, sticky <strong>and</strong> more or less coloured. Its composition<br />
is highly complex (probably thous<strong>and</strong>s of components are present). No wort has ever<br />
been completely analysed. Substances present include simple sugars, dextrins, -glucans,<br />
pentosans, phosphates, dissolved inorganic ions, proteins, peptides <strong>and</strong> amino acids,<br />
nucleic acid breakdown products, lipids, yeast growth factors (vitamins), organic acids,<br />
bases <strong>and</strong> phenolic substances.Sometimes it is desirable toanalyseachemical fraction in<br />
detail, but this is not always the case. A `typical' sweet wort may contain solids<br />
consisting of about 90 92% carbohydrates, 4 5% nitrogen-containing substances <strong>and</strong><br />
Product<br />
t1 t2 t3<br />
Incubation time<br />
Fig. 4.1 Graphs illustrating the appearance of products in idealized enzyme-catalysed reactions<br />
carried out with the initial conditions the same but at different incubation temperatures (ëC) (after<br />
Dixon <strong>and</strong> Webb, 1958). As the temperature is increased so the initial reaction rate, at time zero,<br />
increases. In the sample, at 40 ëC the enzyme is stable <strong>and</strong> so the reaction carries on at a steady rate<br />
(substrate is present in excess), <strong>and</strong> product is formed linearly with time. However, at higher<br />
temperatures the enzyme is less stable <strong>and</strong> so, although the initial reaction rates are more rapid,<br />
enzyme is progressively inactivated <strong>and</strong> the rates of product formation decline. If the maximum<br />
amounts of product formed at different times are noted it can be seen that at the shortest time, t1, the<br />
`optimum' is at 70 ëC, at t2 it is at 60 ëC while at the longest time, t3, it is at 50 ëC.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
50°<br />
60°<br />
70°<br />
40°<br />
80°
1.5 2% ash (MacWilliam, 1968). As yeast ferments wort the simpler sugars are partly<br />
converted into ethyl alcohol <strong>and</strong> carbon dioxide, <strong>and</strong> the specific gravity of the mixture<br />
progressively declines until fermentation is complete. This final value is the attenuation<br />
limit of the wort <strong>and</strong> mainly depends on the carbohydrate spectrum of the wort (Chapters<br />
12 <strong>and</strong> 14). During the hop-boil the fermentability of the wort <strong>and</strong> its strength may be<br />
adjusted by the addition of carbohydrate adjuncts (Chapter 2). Enzymes remain in sweet<br />
wort, slowly increasing its fermentability, which is not fixed until the wort is boiled.<br />
Fermentability values are calculated from the changes in specific gravity.<br />
Wort colour mustbe within specification, <strong>and</strong> so mustthe nitrogen (crude protein ˆN<br />
6.25) fractions of the wort. Total soluble nitrogen (TSN) is self-explanatory. Free<br />
amino nitrogen (FAN) is ameasure of the low molecular weight substances, mainly<br />
amino acids, which are needed to support yeast growth <strong>and</strong> metabolism. Older measures<br />
included coagulable-N (nitrogen containing material that was precipitated when the wort<br />
was boiled) <strong>and</strong> permanently soluble nitrogen, (PSN) which remained in solution.<br />
Formol-N was an older method of estimating the amino acid <strong>and</strong> peptide fraction.<br />
4.2 Mashing schedules<br />
Mashingschedulesvarywidely.Oneischosenwithreferencetothebeertypetobemade,<br />
thewayithasbeenmadeinthepast,theplant<strong>and</strong>rawmaterialsthatareavailable<strong>and</strong>the<br />
energy consumption <strong>and</strong> speed of the process. There is atrend towards temperatureprogrammed<br />
infusion mashing but some brewers have not been able to `match' their<br />
traditional products when changing from older types of mashing programmes, <strong>and</strong> so<br />
thesehavebeenretained.Atpresentarangeofbrewingproceduresareinuse<strong>and</strong>theyare<br />
carried out in many different types of equipment (Chapters 5<strong>and</strong> 6). In many small<br />
breweries it is normal to mash not more than once a day, but in some large production<br />
units the target is to mash 12 14 times every 24 hours. This has necessitated many<br />
changes in equipment <strong>and</strong> mashing <strong>practice</strong>s. It is convenient to distinguish between<br />
traditional infusion mashing, decoction mashing, double mashing, temperatureprogrammed<br />
infusion mashing <strong>and</strong> `all-' or `mainly-adjunct' mashing, although the<br />
distinctions between these classes are not absolute. Some `mixed' mashing systems are<br />
used in Belgium (Briggs, et al., 1981; De Clerck, 1957; Kunze, 1996; Narziss, 1992a;<br />
Hind, 1950; Wright, 1892). The mechanisms of grist preparation <strong>and</strong> mashing are<br />
discussed in Chapters 5 <strong>and</strong> 6.<br />
Before 1945 the infusion mashing carried out in the UK typically involved making a<br />
thick mash with well modified <strong>and</strong> comparatively coarsely ground malt or malts, mixed<br />
with 5 15% of flaked maize or flaked rice adjunct. The grist was mixed with hot liquor<br />
(water) at a temperature (`liquor heat' or `striking heat') chosen to give a particular<br />
`initial heat' or mash temperature. After a st<strong>and</strong> of about 30 minutes, when the mash gave<br />
a negative iodine test for starch, an underlet (hot water added into the bottom of the mash)<br />
might be given to raise the temperature then, after a total period of 2 3 h, wort collection,<br />
recirculation <strong>and</strong> sparging would begin. Typically on mashing in the liquor/grist ratio<br />
would be 2.15 2.42 hl/100kg grist (2 2.25 imp. brl/Qr.) <strong>and</strong> the temperature would be<br />
63.4 67.2 ëC(146 153 ëF). After underletting, with additions of hot water of 0 1.34 hl/<br />
100kg (0 1.25 imp. brl/Qr.) the temperature of the mash would be 66.6 68.8 ëC<br />
(152 156 ëF). Finally, the wort would be collected <strong>and</strong> the goods (residual solids) were<br />
sparged with 3.76 4.30 hl/100 kg (3.5 4 imp. brl/Qr.) of liquor at 75 77 ëC<br />
(167 170.7 ëF). Thus the whole process from mashing in to finishing collecting the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
wort would take at least six hours. In one brewery the whole process took 18 hours. This<br />
process cannot be greatly accelerated.<br />
The choice of good quality malt minimizes the chance of a set mash <strong>and</strong> allows the<br />
st<strong>and</strong> to be shortened to 1 1.5 h, <strong>and</strong> a shortening of the wort separation time <strong>and</strong> the<br />
addition of some hydrolytic enzymes can accelerate wort separation. Mashing directly in<br />
a lauter tun, rather than a mash tun, allows the use of a more finely ground grist <strong>and</strong> faster<br />
wort separation. By shortening the st<strong>and</strong> period <strong>and</strong> by accelerating sparging, time can be<br />
saved but at the risk of reducing extract recovery <strong>and</strong> altering the quality of the wort. The<br />
total volumes of liquor used were 6.98 7.52 hl/100 kg grist (6.5 7.0 imp. brl/Qr.).<br />
Modern infusion mashes are made with 1.6 3.2 hl liquor/100 kg grist (1.5 3.0 imp. brl/<br />
Qr.). Torrefied cereals or wheat flour are commonly used adjuncts. The initial<br />
temperature is usually in the range 63 67 ëC (145.5 152.7 ëF) <strong>and</strong> is best held for<br />
1 3 h. The temperature of the mash rises during sparging. This type of mashing does not<br />
allow air to be excluded from a mash, <strong>and</strong> indeed the entrained air bubbles cause much of<br />
the mash to float. This is no disadvantage, <strong>and</strong> may even be desirable, for making<br />
traditional, cask-conditioned British beers. However, with other beers, intended to have<br />
very long shelf-lives, efforts are increasingly being made to exclude air from the mash<br />
<strong>and</strong> the hot wort. To achieve this equipment other than a mash tun must be used. In<br />
special cases, when alcohol production is to be minimized, the mashing-in temperature is<br />
increased to, e.g., 75 ëC (167 ëF) to allow -amylase to liquefy <strong>and</strong> dextrinize the starch<br />
while minimizing saccharification by -amylase <strong>and</strong> so producing a less fermentable<br />
mixture of carbohydrates.<br />
In traditional continental European decoction mashing a thin mash (3.5 5 hl liquor/<br />
100 kg. grist; 3.26 4.66 imp. brl/Qr.) is made from undermodified malt that is<br />
comparatively finely ground. The thin mash is necessary to permit it to be stirred <strong>and</strong><br />
pumped between mashing vessels. In this, <strong>and</strong> the other mashing systems to be<br />
considered, the mash conversion processes are carried out in vessels that are separate<br />
from the devices (lauter tuns or mash filters) in which the wort is separated from the<br />
residual spent grains. Because the mash is stirred <strong>and</strong> portions of it are pumped between<br />
vessels air is not entrained <strong>and</strong> the solids do not float. When portions of the mash are<br />
boiled the starch is gelatinized <strong>and</strong> becomes susceptible to enzymic attack, residual<br />
cellular structures are disrupted, proteins are denatured <strong>and</strong> precipitated, enzymes are<br />
inactivated, chemical processes are accelerated, flavour substances (not necessarily<br />
desirable) appear in the wort <strong>and</strong> the wort darkens. Unwanted substances such as<br />
pentosans <strong>and</strong> -glucans are extracted. Boiling portions of the mash is expensive because<br />
it involves the consumption of energy. The successive temperatures, which occur in the<br />
`main, mixed mash', allow key enzymes to act at or near their optimal temperatures. In<br />
decoction mashing the grist is mashed into the mash-mixing vessel, which has a stirrer<br />
<strong>and</strong> may have heat-exchanging surfaces to allow the temperature of the contents to be<br />
increased. At intervals aliquots of the mash are withdrawn to the decoction vessel where<br />
they are heated, rapidly or slowly as the programme requires, with or without `rests' at<br />
particular temperatures, to boiling. After a period of boiling the hot material is pumped<br />
back into, <strong>and</strong> is mixed with, the main mash raising its temperature at a predetermined<br />
rate to a pre-chosen value. Before a decoction is carried out the stirrer in the mash-mixing<br />
vessel may be turned off <strong>and</strong> the mash allowed to settle. Then part of the settled `thick<br />
mash' is pumped to the decoction vessel.<br />
If adjuncts are permitted they may be cooked, with some of the malt mash <strong>and</strong><br />
possibly added microbial enzymes, in the decoction vessel. The mash is allowed to st<strong>and</strong><br />
until the next temperature rise, created either by another decoction, or by direct heating or<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
y sparging. At the end of the mash conversion period the mash is transferred either to a<br />
lauter tun or to a mash filter for wort separation. With undermodified malts double<br />
decoction mashing is said to recover 2% more extract than an infusion mash <strong>and</strong> a single<br />
decoction process recovers 1.5% more extract. These gains are made at the expense of<br />
higher energy costs, as boiling part of a mash requires heat. With well-made malts the<br />
advantages, if present, are very small if temperature-programmed infusion mashing is<br />
employed. The brewing problems created by using poorly modified malts are such that<br />
their use is now avoided where possible <strong>and</strong> so the need for decoction mashing is going.<br />
On the other h<strong>and</strong> some mainl<strong>and</strong> European beers have their full <strong>and</strong> desirable range of<br />
characteristics only if their worts are prepared by decoction mashing.<br />
In British infusion mashing, carried out with well modified malts, extract yield is<br />
likely to be limited by the extract recovery from the mash, rather than the extent of the<br />
mash conversion. In other words extract will remain in the spent grains. Decoction<br />
mashing schedules are very flexible <strong>and</strong> are easily adjusted (Kunze, 1996; Narziss,<br />
1992a, b). In the classical three-decoction process (Fig. 4.2) light beers are made with a<br />
liquor/grist ratio of 4.8 5.4 hl/100 kg grist (4.5 5 imp. brl/Qr.) while dark beers are<br />
made with thicker mashes, 3 4 hl/100 kg (2.75 3.75 imp. brl/Qr.). In the decoctions<br />
used when making light beers, the boiling periods are shorter than when dark beers are<br />
being made. The grist may be mashed in with cold water <strong>and</strong> the temperature is raised to<br />
35 40 ëC (95 104 ëF) either by adding hot water, or by direct heating, while the mash is<br />
stirred. The main mash may be allowed to remain at this temperature for about two hours.<br />
During this st<strong>and</strong> heat-labile enzymes, such as -glucanase, maltase, proteases <strong>and</strong><br />
phytase, have a chance to act. The pH of the mash may fall, partly due to the activities of<br />
lactic acid bacteria. After about one hour into this period a third of the mash (stirred `thin'<br />
or settled <strong>and</strong> `thick') is transferred to the decoction vessel <strong>and</strong> is heated to boiling, often<br />
Temperature (°C)<br />
100<br />
75<br />
50<br />
25<br />
1 2 3<br />
Proteolysis active in mash-mixing vessel<br />
Saccharification active<br />
0 32<br />
0 1 2 3 4 5 6<br />
Time (h)<br />
212<br />
167<br />
Enzyme<br />
inactivation<br />
Fig. 4.2 The temperature changes occurring in a typical, traditional triple-decoction mashing<br />
programme (after Hind, 1950). ÐÐÐ , temperatures in the main mash vessel. temperatures<br />
in the mash copper during the first, second <strong>and</strong> third decoctions (1, 2 <strong>and</strong> 3). About one-third of the<br />
mash is used in each decoction. A `thick mash' or a mixed mash may be used. The `proteolysis' <strong>and</strong><br />
other periods are the oversimplified, traditional names for the divisions of the process. Process<br />
duration, about six hours.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
122<br />
77<br />
Temperature (°F)
with arest at 65 70ëC (149 158ëF) to allow -amylase to liquefy the starch. After a<br />
period at 100ëC (212ëF), say 15min. for pale beers <strong>and</strong> 45min. for dark beers, the hot<br />
mixture is added back to the stirred main mash, increasing its temperature to 50 53ëC<br />
(122 127ëF). During the next rest the surviving enzymes begin to attack the gelatinized<br />
<strong>and</strong> liquefied starch <strong>and</strong> proteolysis continues relatively quickly. Asecond decoction,<br />
also with about athird of the mash, which may or may not have a`rest' during heating at<br />
about 65ëC (149ëF) to liquefy starch, increases the temperature of the main mash to<br />
about 65 70ëC (149 158ëF).Afinal decoctionincreasesthe temperature toabout 76ëC<br />
(169ëF) then, after ashort rest, the mash is transferred to the lauter tun or filter <strong>and</strong> wort<br />
collection begins. Athree-decoction mash may last six hours. As with other mashes the<br />
exact temperatures chosen, the duration of the rests <strong>and</strong> boils <strong>and</strong> the rates of mash<br />
heating <strong>and</strong> mixing can be varied. However, with well-modified malts such aprocess is<br />
unnecessary. It is too long, too complex <strong>and</strong> the three boils are too expensive.<br />
Many faster <strong>and</strong> more economical double- <strong>and</strong> single-decoction procedures are used.<br />
For example, malt may be mashed in at 35ëC (95ëF) <strong>and</strong>, after a short rest, the<br />
temperature of the mash is raised to around 52ëC (126ëF; Fig. 4.3). Two successive<br />
decoctions, each with aquarter of the mash, increase the main mash temperature to<br />
65 70ëC (149 158ëF) <strong>and</strong> then 76ëC (169ëF). The whole process takes about 4.5h.<br />
Many morerapid double-decoction processeshave been described<strong>and</strong>ineach case better<br />
modified malts are needed <strong>and</strong> the processes more nearly approach the conditions used in<br />
infusion mashing. For example, amash is prepared at 63ëC (145ëF) <strong>and</strong> after ashort<br />
st<strong>and</strong> about aquarter of the mash, possibly a`thick mash', is withdrawn <strong>and</strong> boiled (Fig.<br />
4.4). When it is mixed into the main mash the temperature is increased to 70ëC (158ëF).<br />
After arest of 45 60min. asecond decoction, also with about aquarter of the mash, is<br />
used to increase the main mash temperature to about 77ëC (171ëF). After about 30min.<br />
wort collection can begin. The whole process takes 2 3h.<br />
Single decoction mashes are even simpler. For the preparation of dark beers the mash<br />
may be given a preliminary long st<strong>and</strong> at a low temperature. This can allow the<br />
Temperature (°C)<br />
100<br />
75<br />
50<br />
25<br />
Proteolysis active<br />
1 2<br />
Saccharification active<br />
0 0 1 2 3 4<br />
Time (h)<br />
Enzyme<br />
inactivation<br />
Fig. 4.3 Atemperature scheme for atypical two-decoction mashing programme (after Hind,<br />
1950). The key is in Fig. 4.2. About aquarter of the mash is used in each decoction. Process<br />
duration, four <strong>and</strong> a half hours.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
212<br />
167<br />
122<br />
77<br />
32<br />
Temperature (°F)
Temperature (°C)<br />
100<br />
75<br />
50<br />
25<br />
Proteolysis<br />
1 2<br />
Saccharification active<br />
0 32<br />
0 1 2 3<br />
Time (h)<br />
Enzyme<br />
inactivation<br />
Fig. 4.4 Ashortened double-decoction process, lasting about two <strong>and</strong> ahalf hours (after Hind,<br />
1950). For the key, see Fig. 4.2. About aquarter of the mash is used in each decoction.<br />
multiplication of unwanted microbes <strong>and</strong> the development of unwanted flavours. Often<br />
the single decoction is coupled with atemperature-programmed period. For example, a<br />
mash is made at 35ëC (95ëF), then the temperature is successively raised to 50ëC<br />
(122ëF)<strong>and</strong>then65 67ëC(149 152.7ëF)withrestsatthesetemperatures.Thenabouta<br />
third of the mash (a thick mash) is taken <strong>and</strong> heated, with arest at 70ëC (158ëF), to<br />
boiling. This is added back to the main mash, increasing the temperature to 75ëC<br />
(167ëF). An interesting variant is where amash is made at 35 or 50ëC (95 or 122ëF) <strong>and</strong><br />
after ast<strong>and</strong> the mash is allowed to settle <strong>and</strong> the relatively clear liquid, which contains<br />
enzymes,isheldwhilethethickmashisstirred<strong>and</strong>directlyheatedwithrestsat63 65ëC<br />
(145 149ëF) <strong>and</strong> 70 75ëC (158 167ëF), then the thick mash is heated to boiling. This<br />
boil will disrupt any solids <strong>and</strong> gelatinize any remaining starch granules but will<br />
inactivate enzymes. Then the thick mash is cooled to 65ëC (149ëF) <strong>and</strong> is recombined<br />
with the thin mash, which provides enzymes to attack the disrupted materials. The<br />
recombined mash is initially at about 67ëC (152.7ëF), <strong>and</strong> is successively warmed, by<br />
direct heating, to 70 <strong>and</strong> then 75ëC (158 <strong>and</strong> 167ëF). However, the cooling process<br />
wastes heat.<br />
Various special mashing programmes are used in Germany (Kunze, 1996; Narziss,<br />
1992a,b). In the jump-mash system (Springmaischverfahren), which is used to produce<br />
wort with alow fermentability, athick mash is prepared at 35 40ëC (95 104ëF). Then<br />
boiling water is stirred in over a15min. period to give atemperature of 72ëC (161.6ëF).<br />
By this means the grist is hydrated <strong>and</strong> some of the thermolabile enzymes have achance<br />
to act before the temperature is increased to permit starch liquefaction <strong>and</strong> dextrinization<br />
while minimizing saccharification. The mash temperature is increased to about 78ëC<br />
(172.4ëF) before wort collection. The wort has an attenuation limit of only about 40%. In<br />
the Kubessa process the grist is divided into flour, grits <strong>and</strong> husk fractions. The husk<br />
fraction is mashed separately at 50ëC (122ëF) <strong>and</strong> is held at this temperature while the<br />
flour <strong>and</strong> grits are mashed using a rising temperature programme, with rests at<br />
appropriate temperatures, until the mix is boiled. Then the two mashes are combined to<br />
giveamixed mashatabout70ëC(158ëF).Afterast<strong>and</strong>thetemperatureisraised to78ëC<br />
(172.4ëF) <strong>and</strong> the wort is collected. This process, which is little used, avoids boiling the<br />
huskmaterial <strong>and</strong> gives beer with abetter flavour. Inthe preparation oflow-carbohydrate<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
212<br />
167<br />
122<br />
77<br />
Temperature (°F)
eers it is necessary to ferment as much of the wort carbohydrate as possible. Where the<br />
use of microbial enzymes is permitted this is achieved by adding fungal -amylase, with<br />
or without pullulanase, to the fermenting beer. In Germany highly fermentable wort may<br />
be made by using an exceptionally `intensive' temperature-programmed mash, with rests<br />
at 50 ëC (122 ëF)/30 min.; 62 ëC (143.6 ëF)/45 min.; 65 ëC (149 ëF)/45 min.; 68 ëC<br />
(154.4 ëF)/30 min.; 70 ëC (158 ëF)/30 min.; 72 ëC(161.6 ëF)/15 min. <strong>and</strong> then mashing off<br />
at 73 74 ëC (163.4 165.2 ëF). This process takes 3.5 4 h. Even so the wort is not fully<br />
fermentable <strong>and</strong> it is necessary to add powdered highly diastatic malt or malt extract to<br />
the fermenting beer. These additions risk contaminating the beer with spoilage<br />
organisms.<br />
Decoction mashing is convenient when small amounts of adjuncts need to be cooked,<br />
since cooking can be carried out in the decoction vessel. The use of large quantities of<br />
adjuncts, such as rice, maize <strong>and</strong> sorghum grits, that require thorough cooking, combined<br />
with the availability of high-nitrogen, enzyme-rich malts gave rise, initially in North<br />
America, to the double-mash system. With this two mashes are prepared <strong>and</strong> then they<br />
are combined, often in a third vessel (Fig. 4.5). The adjuncts, which may comprise<br />
25 60% of the grist, are mashed in, in a cereal cooker, with a proportion (5 10%) of a<br />
highly enzymic malt (80 200 ëL) or a heat-stable bacterial -amylase, at about 35 ëC<br />
(95 ëF). The temperature is raised to about 70 ëC (158 ëF) <strong>and</strong> the malt starch is liquefied<br />
<strong>and</strong> liquefaction of the adjunct starch begins. The temperature is traditionally increased to<br />
boiling <strong>and</strong> is held at this temperature for about 45 min. However, depending on the grist<br />
<strong>and</strong> the enzymes used, it may be preferable to hold the temperature at 85 ëC (185 ëF), <strong>and</strong><br />
so save the cost of the fuel needed for boiling. While the adjunct mash is in progress the<br />
malt mash is prepared by mashing-in (doughing-in) at about 35 ëC (95 ëF). After a rest of<br />
about one hour the two mashes are combined <strong>and</strong> mixed achieving a temperature of about<br />
68 ëC (154.4 ëF). After a st<strong>and</strong> of 15 30 min., when all the starch is saccharified, the<br />
mash is heated to around 73 ëC (163 ëF) by adding hot liquor or by steam injection, then it<br />
is lautered.<br />
Temperature (°C)<br />
100<br />
75<br />
50<br />
25<br />
Saccharification<br />
in first mash<br />
Saccharification<br />
in second mash<br />
0 32<br />
0 1 2 3<br />
Time (h)<br />
Enzyme<br />
inactivation<br />
Fig. 4.5 A scheme of the temperatures found in a double-mash programme using a grits cooker<br />
(after Hind, 1950). the temperature of the adjuncts/grits mash in the cereal cooker. ÐÐÐ the<br />
temperature in the malt mash <strong>and</strong> in the combined mash during <strong>and</strong> after mixing.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
212<br />
167<br />
122<br />
77<br />
Temperature (°F)
In parts of Africa no barley malt is available <strong>and</strong> the sorghum malt is not suitable for<br />
making lager-types of beers with mashing schedules designed for barley malts (Chapter<br />
16). Under these circumstances mashes have been made with maize or sorghum grits<br />
converted with microbial enzymes. Sometimes 10 20% of sorghum malt has been<br />
included in these grists to provide soluble nitrogenous materials, but they are virtually<br />
`all-adjunct' mashes.<br />
When making some traditional Belgian top-fermented beers the temperature of the<br />
infusion mash may be increased by steam injection (De Clerck, 1957). Mashing-in may<br />
be at 45 50 ëC (113 122 ëF) then, after a 30 45 min. st<strong>and</strong> the temperature is increased<br />
to 62 63 ëC (143.6 145.4 ëF). After another rest of 30 45 min. the temperature is raised<br />
to 70 ëC (158 ëF) <strong>and</strong> then, after 30 45 min., to 75 ëC (167 ëF). After a pause wort<br />
collection is started. Temperature-programmed infusion mashing is being more widely<br />
used both in ale <strong>and</strong> lager breweries. The rising temperature programmes may be adjusted<br />
in many ways, allowing rests at any desired temperature. The stirred mash is heated in<br />
one vessel, sometimes with precautions to exclude air, <strong>and</strong> so the costs of a decoction<br />
vessel <strong>and</strong> heating parts of the mash to boiling are avoided, although an adjunct cooker<br />
may be needed. Boiling thick mashes is not practical <strong>and</strong> so, if undermodified malts or<br />
mashes with particular adjuncts are used, the mashing programmes must be extended <strong>and</strong><br />
it is sometimes necessary to add microbial enzymes. For a mash made with an<br />
undermodified malt the temperature/time sequence might be 35 ëC (95 ëF)/30 min.; 50 ëC<br />
(122 ëF)/30 min.; 65 ëC (149 ëF)/30 min.; 70 ëC (158 ëF)/30 min., 75 ëC (167 ëF)/15 min.<br />
then mashing off, with the temperature rising between the rests at the rate of 1 ëC (1.8 ëF)/<br />
min. For mashes being made with better modified malts the programme might start at<br />
48 50 ëC (118.4 122 ëF) <strong>and</strong> the durations of the different rests may be varied to<br />
achieve the desired quality of wort. Typically these mashes last 2 3 h. Often the mashing<br />
programme is chosen to produce a wort that is closely similar to one made by a decoction<br />
mashing programme (Hug <strong>and</strong> Pfenninger, 1979; Fig. 4.6). Temperature-programmed<br />
Temperature<br />
134.6°F<br />
125.6°F<br />
57°C<br />
52°C<br />
52°C<br />
52°C<br />
57°C<br />
72°C<br />
72°C<br />
62°C<br />
143.6°F<br />
62°C<br />
62°C<br />
168.8°F<br />
161.6°F<br />
76°C<br />
72°C<br />
100°C<br />
72°C 76°C<br />
100°C 100°C<br />
72°C<br />
76°C<br />
0 20 40 60 80 100 120 140 160 180<br />
Time (min.)<br />
Fig. 4.6 The temperature programmes of three mashes that yield very similar worts (after Hug <strong>and</strong><br />
Pfenninger, 1979). The uppermost scheme is for a temperature-programmed infusion mash while<br />
the central scheme is for a single-decoction mash <strong>and</strong> the lowest is for a double-decoction mash.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mashing is easily automated <strong>and</strong> is said to use 30 50% less energy than asimilar<br />
decoction mashing programme. However, while wort produced by different programmes<br />
may often be matched, this is not always the case. Decoction mashes tend to give darker<br />
worts with lower TSN levels <strong>and</strong> higher viscosities due to the non-starch polysaccharides<br />
dissolved during boiling. The husks are not boiled in infusion mashing. This is said to<br />
improve the flavour of the resultant beer.<br />
Novel `mixed' mashing systems are used in Belgium for preparing some traditional<br />
beers (De Clerck, 1957). These involve the use of large amounts of unmalted wheat. In<br />
one system the wheat is boiled in acopper, <strong>and</strong> then cooled <strong>and</strong> adiastatic extract is<br />
added to liquefy the starch. Separately the malt is mashed in amash tun <strong>and</strong> the wort is<br />
drawn off from the base. The wheat mash is transferred onto the top of the malt mash <strong>and</strong><br />
the liquid is collected after it has filtered through the layer of malt. Other mashing<br />
systems involve producing `turbid worts' by mashing malt <strong>and</strong> wheat, mixed together, in<br />
a mash tun at 50ëC (122ëF), st<strong>and</strong>ing <strong>and</strong> then collecting the wort <strong>and</strong> the turbid<br />
supernatant in acopper. The re-mashing is repeated two or three more times, using<br />
progressively hotter liquor. Each time the wort is added to the copper. The copper<br />
contents are heated <strong>and</strong> held at 70ëC (158ëF) to saccharify suspended starch. This<br />
process, which produces turbid final worts, is needed to give Lambic beer its correct<br />
character.<br />
4.3 Altering mashing conditions<br />
4.3.1 The grist<br />
Malt<strong>and</strong>someadjuncts,suchastorrifiedwheat,mustbebrokenupbeforemashing.Until<br />
recently this was nearly always achieved by roller milling, but now hammer-milling is<br />
sometimes used (Chapter 5). The objective of milling is to break up the grist to give an<br />
acceptable range of particle sizes. The acceptable range is determined by the wort<br />
separation system being used. Often milling is carried out in such away as to minimize<br />
the break-up of husk material, as husk fragments help to give the mash an open structure<br />
<strong>and</strong>aidwortseparation.Indeed,manyyearsagooat huskswere addedtomashesto`open<br />
them up'. Mash tuns require the coarsest grists, followed by various types of lauter tun,<br />
<strong>and</strong> then mash filters. While older types of mash filters required afine, roller-milled grist<br />
the newest designs use hammer-milled grists that are very fine indeed. There are reasons<br />
for using the most finely ground grist that can be processed with the equipment available.<br />
The finer the particles the faster they hydrate on mashing, the faster the pre-formed<br />
soluble substances dissolve <strong>and</strong> the faster the extract leaches from the particles during<br />
sparging. Furthermore, the enzymes have more ready access to their substrates in<br />
thoroughly disrupted grists. The surface/volume ratio of amaterial is larger the smaller<br />
<strong>and</strong> more numerous the particles into which it is divided <strong>and</strong> so afinely divided grist<br />
provides alarger surface area on which enzymes can act <strong>and</strong> across which substances can<br />
diffuse. To varying extents the particles will be pervious, permitting enzymes, substrates<br />
<strong>and</strong> the products of hydrolysis to enter <strong>and</strong> leave. In <strong>practice</strong>, finer grinding gives grists<br />
that, up to agiven `degree of fineness', yield higher extracts (Tables 4.1, 4.2). Finer<br />
grinding is less advantageous with better modified malts (this is the basis of the analytical<br />
fine-coarse extract difference determination), but it is beneficial with many (perhaps all)<br />
mash tun adjuncts.<br />
For many years attempts have been made to find ways of producing worts from finely<br />
ground, or `pulverized' grists. Decanter centrifuges, belt filters <strong>and</strong> rotary vacuum filters<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.1 Average laboratory extract values, obtained from 12 malts mashed isothermally at<br />
65 ëC (159 ëF) for one hour, using three settings of the BuÈhler-Miag disc mill. The smaller the gap<br />
between the grinding surfaces the finer the grind (Martin, 1979)<br />
Gaps between milling surfaces (mm) 0.2 0.5 0.7<br />
Hot water extract (l ë/kg) 296.2 293.5 291.4<br />
Confidence limits (95%) 2.0 1.9 1.9<br />
(In a previous trial the HWE values were 305.4, 302.3 <strong>and</strong> 300.0 respectively.)<br />
Table 4.2 Some analyses of two short-grown, experimental malts <strong>and</strong> a commercial malt ground<br />
with different degrees of fineness (Wackerbauer et al., 1993)<br />
Malts (days germination) 3 5 7 (Commercial)<br />
Friability (%) 62 83 91<br />
Whole corns (friabilimeter, %) 3 2 1<br />
Extract (coarse grind EBC,%) 73.9 78.6 79.7<br />
Extract (fine grind EBC, %) 80.2 80.8 80.9<br />
Extract (hammer milled, %) [78.6]* 81.4 81.7<br />
* The reason for this atypical low value is not clear. Possibly the mixing in of the very fine grist was uneven.<br />
(used with kieselguhr filter aids) have been tried for separating the wort from the spent<br />
grains.Theadventofthenewesttypesofmashfilters(Chapter6)haspermittedtheuseof<br />
very fine grists. The introduction of very fine grinds necessitates the alteration of the<br />
mashing schedule if an established product is to be `matched'. More finely ground grists<br />
are `converted' more quickly, saccharify faster, give higher extracts <strong>and</strong> sometimes the<br />
worts obtained are more fermentable <strong>and</strong> less turbid. Narziss (1992b) reported that malts<br />
ground coarsely, finely <strong>and</strong> powdered gave samples with extracts (% dry wt.) of 78.9,<br />
80.7 <strong>and</strong> 82.4, the worts having real attenuation limits (%) of 66.1, 65.1 <strong>and</strong> 65.3<br />
respectively. By using very fine grists the whole mashing process can be carried out more<br />
quickly <strong>and</strong> with better extract yields. The levels of TSN <strong>and</strong> FAN increase <strong>and</strong>, at least<br />
with some grists, the levels of soluble -glucans increase <strong>and</strong> wort viscosities increase<br />
(Pollock <strong>and</strong> Pool, 1968; Narziss, 1992a, b; Kunze, 1996). Sometimes the flavour of the<br />
beer produced is improved, perhaps because the shortened mashing times allow less<br />
poorly flavoured material to be extracted. It may be possible to use a higher proportion of<br />
adjuncts in a finely divided grist, <strong>and</strong> the fine grind is helpful when processing undermodified<br />
malts. With modern mash filters the extract recoveries from finely ground grists<br />
can equal, or exceed, laboratory extracts.<br />
Malt grists can be fractionated by sieving (screening) <strong>and</strong>/or by air classification. The<br />
fractions have different compositions <strong>and</strong> yield different worts when mashed separately.<br />
So, for example, fine flour derived mainly from the inner starchy endosperm saccharifies<br />
well when mashed alone, <strong>and</strong> yields an exceptionally high extract (e.g. 96%), <strong>and</strong> gives<br />
pale beers with very fine, pure <strong>and</strong> fresh flavours but lacking in body, low in phenolic<br />
tannins <strong>and</strong> resistant to the formation of chill haze. In contrast the fraction enriched with<br />
the outer parts of the starchy endosperm yields more soluble nitrogen, has an extract of<br />
about 80%, an intermediate colour <strong>and</strong> gives a beer that is full bodied <strong>and</strong> with a fresh<br />
flavour but having a harsh, clinging, astringent or bitter after-taste (Kieninger, 1969,<br />
1972). The husk fraction was obtained in smaller amounts. By combining fractions in<br />
different proportions different types of beer could be made. Also, by removing part or all<br />
of the husk material, brewing with the remainder of the grist would be more rapid, the<br />
beers produced would be paler <strong>and</strong> have a higher haze stability <strong>and</strong> a `finer' flavour<br />
(Vose, 1979). These fractionation processes take time <strong>and</strong> involves extra costs <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
produce ahusk-enriched fraction that needs to be used. Its use in cattle food would seem<br />
to be uneconomic <strong>and</strong> to use it, added to the grist, in the production of `normal lagers',<br />
would compromise their quality. Probably grist fractionation processes are not in<br />
commercial use.<br />
By using ahigh-impact mill, running at areduced speed, <strong>and</strong> then fractionating the<br />
gristovera1.60mmscreenKrottenthaleretal.,(1999)separated thegrist,whichinitially<br />
had an unusually wide range of particle sizes, into 79% finely ground <strong>and</strong> 21% more<br />
coarsely ground. The husk contents of the fractions seemed to be the same, <strong>and</strong> the<br />
extracts were almost identical. However, on analysis the coarse fraction took 55 60min.<br />
tosaccharify,comparedtothe10min.bythefinefraction,<strong>and</strong>thecoarsefractionyielded<br />
less FAN <strong>and</strong> total soluble nitrogen but substantially more -glucan <strong>and</strong> amore viscous<br />
wort. Thus the coarse material was derived from the under-modified portions of the malt.<br />
By mashing in the coarse material at 35ëC (95ëF) then temperature programming to<br />
65ëC (149ëF) then, after arest during which starch conversion should have occurred,<br />
combining the `coarse mash' with the mash of the fine fraction, that had been made at<br />
25ëC (77ëF) in asecond vessel, acombined mash temperature of 45ëC (113ëF) was<br />
obtained. After arest, during which the temperature-sensitive -glucanase surviving in<br />
the fine fraction mash should have operated, the temperature of the combined mash was<br />
increasedto65ëC(149ëF).Then,afterafurtherrestduringwhichthestarchfromthefine<br />
fraction should have been converted, the temperature was increased to 72ëC (161.6ëF),<br />
which was held for afurther period, allowing some final -amylolysis, before lautering.<br />
Comparedtothest<strong>and</strong>ardmashingprogramme,slightlymoreextractwasrecoveredusing<br />
the fractionated grist mash (83.0 compared to 83.5%) the viscosity of the wort was<br />
usefully reduced (1.61 compared to 1.52 mPa.s for an 8.6% wort) <strong>and</strong> the -glucan<br />
content wasroughlyhalved.Suchatechniquemay beattractivewheretheReinheitsgebot<br />
or similar restrictions are in force. Asimilar result could be obtained more simply by<br />
adding afungal -glucanase/cellulase preparation to the unfractionated mash.<br />
4.3.2 Malts in mashing<br />
The choice of malts is dictated by the type of beer to be made <strong>and</strong> by quality<br />
considerations. Some qualities of different types of malt are indicated in Chapter 2. The<br />
brewer is faced with the problem that malts with the same traditional analyses may be<br />
different, <strong>and</strong> the differences can give rise to major problems in the brewery <strong>and</strong> in beer<br />
quality. Coloured <strong>and</strong> special malts' flavours change <strong>and</strong> decline with age <strong>and</strong> so these<br />
materials should be used fresh <strong>and</strong> their lab worts should be tasted <strong>and</strong> smelled to see that<br />
they are `normal'. Although chemical `marker' substances, such heterocyclic, nitrogencontaining<br />
Maillard products, have been sought, to allow flavour to be quantified<br />
indirectly by chemical analyses, this approach has had little success. Most attention has<br />
been paid to pale malts, since these make up the greater part of malt requirements,<br />
whether or not adjuncts are also used. In the first place each batch of malt should be as<br />
nearly identical as possible to earlier batches of the same type used successfully to make<br />
a particular product. This is not necessarily easily achieved. The old belief that some pale<br />
malts behave better when mashed after several weeks storage has been confirmed<br />
(Rennie <strong>and</strong> Ball, 1979). The ease of wort separation improves over a period of about<br />
three weeks <strong>and</strong> the clarity of the worts improves, but the reason(s) are unknown.<br />
As the available varieties of barley change the problem arises that they produce<br />
different malts when malted in one way. Some varieties give malts that give higher, or<br />
lower levels of hydrolytic enzymes, TSN or FAN relative to the yield of extract, or that<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
give worts differing in fermentability or flavour <strong>and</strong> so on. Even comparatively small<br />
differences between samples of one variety of barley can cause differences during<br />
malting <strong>and</strong> in the quality of the malt produced. Irregularities in germination can lead to<br />
inhomogeneity, which is not always easy to detect, <strong>and</strong> may not be suspected until it has<br />
caused problems in the brewery.<br />
Brewers require amalt that mills easily to give the correct rangesof particle sizes, that<br />
converts in the `st<strong>and</strong>ard' time to consistently produce their st<strong>and</strong>ard wort, <strong>and</strong> allows<br />
easy, rapid <strong>and</strong> repeatable wort separation from the mash. The recovery of the extract<br />
should be as high as the equipment in use allows. These points are particularly important<br />
in automated plant<strong>and</strong>/orwhereahigh, fixed number of brews should becompleted each<br />
day <strong>and</strong> there is no spare time. The wort should have the correct characteristics for the<br />
beer being made <strong>and</strong> the beer made from it should be easy to filter <strong>and</strong> require minimum<br />
`stabilization treatments' (e.g. with silica hydrogel, PVPP adsorbents or additions of<br />
enzymes) to minimize haze formation or flavour deterioration <strong>and</strong> so have amaximum<br />
shelf-life.Thewortshouldhaveallthecomponents theyeastneedstoachieve arapid<strong>and</strong><br />
complete fermentation.<br />
Brewers require analyses of each batch of malt. Regrettably different brewers have<br />
different requirements (Chapter 2). Normally analyses will include moisture content,<br />
colour, laboratory extract (HWE or E), total nitrogen (TN; protein), soluble nitrogen (or<br />
protein) <strong>and</strong> free amino nitrogen (FAN, determined by astated method). Sometimes, <strong>and</strong><br />
often if adjuncts are being used, they will also ask for estimates of the diastatic power<br />
(DP) <strong>and</strong> -amylase activity. -Glucanase estimations may also be required. These are of<br />
minimum value if isothermal infusion mashing is being used unless barley adjuncts are<br />
included in the grist. However, they can be of use when deciding on the decoction or<br />
temperature-programmed mashing sequence to be used with under-modified malts or<br />
grists containing barley adjuncts.<br />
The most important analysis is the laboratory extract as, in general, the higher this is<br />
the better the quality of the malt. Allowance is made for the nitrogen content, which is<br />
inversely related to the extract yield, <strong>and</strong> for the variety of barley from which the malt is<br />
made. High total nitrogen contents are related to better foam characteristics in the beer<br />
<strong>and</strong>tohardermalts.AnadequateFANlevelisneededinaworttoensurethatyeastgrows<br />
well <strong>and</strong> that fermentation proceeds rapidly <strong>and</strong> completely. Too high values are not<br />
desired as this can lead to excessive <strong>and</strong> wasteful yeast growth. Higher values are<br />
required for malts that are going to be used together with adjuncts that act as nitrogen<br />
diluents as they contribute relatively little or no soluble nitrogen to the wort. Colour, or<br />
boiled wort colour, gives agood estimate of the colour of the beer <strong>and</strong> indicates what<br />
colour adjustment may be needed.<br />
Other tests are now often used in attempts to overcome the deficiencies of the<br />
traditional analyses (Briggs, 1998). Heterogeneity may be determined by scoring stained<br />
grain sections or by the determination of partially unmodified grains using the<br />
friabilimeter. The friability of the grain indicates the type of result that will be obtained<br />
when the malt is milled. Wort -glucan may be determined as may the residual -glucan<br />
in the malt <strong>and</strong> the viscosity of the wort. High values for fine grind-coarse grind<br />
laboratory extract differences indicate that malts are under-modified <strong>and</strong> that they may<br />
give rise to brewing problems (Table 4.2). At least one brewery has found that the fine<br />
grind, coarse grind <strong>and</strong> concentrated mash extract difference <strong>and</strong> the total malt -glucan<br />
content are inversely related to extract recovery in the brewhouse <strong>and</strong> that the viscosity of<br />
wort from a 70 ëC laboratory mash is correlated with the viscosity of strong brewery<br />
worts <strong>and</strong> so high values give warning of possible beer filtration problems <strong>and</strong> the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
occurrence of -glucan hazes <strong>and</strong> gels (Bourne <strong>and</strong> Wheeler, 1982, 1984; Bourne et al.,<br />
1982). In another brewery, using traditional ale isothermal infusion mashing, the extract<br />
recoveries in the brewery were correlated with the hot water extracts <strong>and</strong> fine-grindcoarse-grind<br />
extract differences. So, from the laboratory measurements it was possible,<br />
by entering the values in the appropriate equation, to predict the brewhouse yield of<br />
extract (Maule <strong>and</strong> Crabb, 1980). Many correlations between laboratory analyses <strong>and</strong><br />
brewery performances have been reported, but the correlation coefficients seem to vary<br />
significantly or tofail in different years <strong>and</strong>/ornot to be applicable todifferent breweries.<br />
Thus the performance of each brewing line needs to be evaluated <strong>and</strong> ways to predict its<br />
performance need to be assessed individually.<br />
`Problem malts', malts which give rise to brewing difficulties, are usually<br />
characterized by their `wrong' degrees of modification, either in all the grains or in a<br />
proportion of the grains, when the malt is inhomogeneous. As noted, the `correct' malt<br />
characteristics vary with the way it is to be used. In addition to inadequate enzyme levels<br />
under-modified malts are characterized by the inadequate breakdown of the endosperm<br />
cell walls. These unmodified regions resemble raw barley, <strong>and</strong> the problems associated<br />
with their presence resemble the difficulties encountered when raw barley is used as an<br />
adjunct. They are tough <strong>and</strong>, when the malt is milled, they give rise to coarse grits. The<br />
intact cell walls contain -glucan <strong>and</strong> pentosans <strong>and</strong>, as they `box in' the starch <strong>and</strong><br />
protein of the cells, these are not degraded because enzymes cannot pass through the cell<br />
walls except where these are disrupted by milling or heating, as in decoctions. So<br />
undermodified malts give low extract recoveries, <strong>and</strong> the worts are often poorly<br />
fermentable. The levels of soluble nitrogen are low, the worts are viscous <strong>and</strong> rich in -<br />
glucans <strong>and</strong> wort run-off is slow. The -glucans may, or may not, deposit as gels or give<br />
rise to hazes, but they always seem to give difficulties with beer filtration. These<br />
problems can be minimized by adding microbial enzymes to the mash. In addition the<br />
beers may show protein-polyphenol haze <strong>and</strong> flavour instabilities. Nor are over-modified<br />
malts desirable. Besides the high malting losses accumulated during their production<br />
these give rise to too powdery grists when the malts are milled (impeding wort run-off)<br />
<strong>and</strong> although the yield of extract is good the quality can be poor. In particular the beer<br />
made from this malt is likely to lack body, the flavour may be poor, <strong>and</strong> the foaming<br />
characteristics will be bad.<br />
Different types of malt have different characteristics. For example, the more highly<br />
cured amalt is the lower its enzyme content (Table 4.3). The extract is slightly reduced<br />
by more curing, <strong>and</strong> the levels of soluble nitrogen are reduced, as shown by the decline<br />
in the nitrogen index of modification. S-Methyl methionine, SMM, the precursor of<br />
dimethyl sulphide, DMS, is destroyed <strong>and</strong> so kilning may be used to regulate the levels<br />
of this compound. The fermentability of the wort is reduced, giving beer with a lower<br />
alcohol content <strong>and</strong> more residual carbohydrate. The colour of the worts from more<br />
highly kilned malts are darker. Crystal malts <strong>and</strong> black malts are enzyme free <strong>and</strong> their<br />
inclusion in a mash reduces the fermentability of the wort. In making low-alcohol beers<br />
it is usual to mash well-cured malts with caramel malts at high temperatures to minimize<br />
saccharification, <strong>and</strong> so reduce the production of fermentable sugars. In addition<br />
experiments have been made in steaming green malts to cause enzyme destruction<br />
before kilning (Briggs, 1998). Curiously, the use of malts dried at low temperatures<br />
(40 ëC, 104 ëF) to a moisture content of 7 8%, which have high enzyme contents, seems<br />
not to occur although the reduction in kilning costs should make them less expensive.<br />
The use of undried, `green' malts is impractical in production brewing. Malts made with<br />
barleys containing a mutation that prevents the formation of anthocyanogen<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.3 Some effects of malt kilning on wort <strong>and</strong> beer analyses. The green malt was freeze<br />
dried before analysis. The kiln-dried samples were removed at successive stages of kilning (data of<br />
MacWilliam, 1972)<br />
Malt properties<br />
Freeze-dried Lightly Kilned Strongly<br />
kilned kilned<br />
Colour ± 3 6 13<br />
Moisture (%) 41.3* 3.7 2.9 2.3<br />
Hot water extract (lb/Qr) 104.0 102.9 102.2 101.9<br />
(l ë/kg) y 308 305 303 302<br />
Cold water extract (%) 19.4 19.5 19.1 17.1<br />
Diastatic power ( ëL) 131 98 68 48<br />
Total nitrogen content (%) 1.59 1.55 1.56 1.47<br />
Index of nitrogen modification (%) z 43.7 43.9 41.6 38.7<br />
[* Moisture content before freeze drying. y Approximate equivalents to the values in the older units. z PSN/TN].<br />
Beer properties<br />
Total carbohydrate (g/l) 13.3 16.7 21.6 27.4<br />
Residual fermentable sugars (g/l) 1.1 2.8 1.3 4.8<br />
Non-fermentable carbohydrate (g/l) 12.2 13.9 20.3 22.6<br />
Total soluble nitrogen (mg/l) 526 645 593 580<br />
Bitterness (EBC units) 22.3 25.8 22.6 23.6<br />
Head retention (half-life, seconds) 91 106 93 101<br />
Colour (EBC units) 12 15 17 27<br />
Flavour (bottled) `Green malt' `Lager-like' `Pale ale' `Mild ale'<br />
polyphenols, which contribute to the formation of protein-tannin hazes, give beers that<br />
are extremely resistant to haze formation, are in limited use. However, proposals to<br />
make malts from low -glucan barley mutants or barleys that have been genetically<br />
modified to contain more heat-stable -amylase or -glucanase have not been carried<br />
out. In part this may be due to the sentiment opposing the use of genetically modified<br />
materials in brewing.<br />
Several brewing problems are associated with microbial infections of malts. Offflavours<br />
may occur <strong>and</strong> there is always the concern that mycotoxins may be present on<br />
poor malts. Particular attention has been paid to the possible presence of aflatoxins,<br />
ochratoxin, zearalenone, deoxynivalenol, fuminosins <strong>and</strong> citrinin, which can be produced<br />
by a range of fungi infecting barley (Scott, 1996). Some, such as citrinin, do not survive<br />
the brewing process, but others, such as deoxynivalenol, can survive into beer. Fungi also<br />
produce factors that cause gushing (over-foaming) in beers. The solution seems to be to<br />
avoid making malts from cereals that are heavily infected with fungi. Fungal infections<br />
are a considerable problem in tropical areas. High levels of bacteria on malt can also give<br />
problems. Bacteria multiply very greatly during malting, especially on the substances<br />
leached from split grains. Malts made with heavily infested barleys have, on mashing,<br />
given rise to very slow wort filtrations, possibly due to microbe-produced polysaccharides<br />
clogging the grain bed. Other malts have given worts having persistent hazes due to<br />
suspended dead bacteria, about 0.6 m in diameter (Walker et al., 1997). Another problem<br />
caused by microbial infestations of malts are the wild, <strong>and</strong> unpredictable fluctuations in<br />
the pH of worts (e.g. pH 5.45±6.06; Stars et al., 1993). Multiplication of lactic acid<br />
bacteria on the growing malt <strong>and</strong> particularly in the initial stages of kilning high-moisture<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
green malts, was largely to blame. The malting process had to be modified to minimize<br />
this problem.<br />
The most uniform homogeneous malts are made by malting grain of one variety <strong>and</strong><br />
onegrade.However,successivebatchesofmalt,madetomeetonespecificationinevitably<br />
differ slightly <strong>and</strong> so they may be mixed, or `blended', to meet abrewer's specification.<br />
Many European brewers regard blending malts of one grade but made from different<br />
barleys as unacceptable, even though in some other areas mixtures of barleys are malted.<br />
Brewers commonly mix different malts (pale, caramel, brown, etc.) to obtain the mix<br />
appropriate for aparticular beer. Narziss (1991, 1992a) gives examples of malt mixtures<br />
used to make many European types of beers. At the start of anew season the old season's<br />
malt, of any type, will increasingly be diluted with the new season's malt of the same type<br />
so that any consequent small differences in beer quality will not be noticed. Some types of<br />
blending are never acceptable. For example, suppose abeer is made with acoloured malt<br />
to give acolour 10. If the usual coloured malt is not available it is not acceptable to blend<br />
50:50 two malts of colours 5<strong>and</strong> 15. The colours may match in intensity (but probably<br />
not quality) but the flavour of the product will not match the original, since the `average'<br />
mix of flavour substances will be different from that in the original malt. When two<br />
similarmaltsareblendeditisnecessarytobeabletopredictthewortqualitythattheblend<br />
will give (Moll, et al., 1982; Yamada <strong>and</strong> Yoshida, 1976). In general the extract <strong>and</strong> FAN<br />
values of mixtures vary linearly between the values of the individual malts according to<br />
their proportions in the mixures. But the fermentabilities of the worts will be better than<br />
predicted from simple proportions (synergism is shown) as enzymes from the more<br />
enzyme-rich malt partly compensate for the inadequate levels in the other. Since in one<br />
case the increase above the fermentability was due to increased levels of maltose <strong>and</strong><br />
maltotriose this could have been due to the activity of limit dextrinase.<br />
4.3.3 Mashing with adjuncts<br />
The characteristics of commonly used adjuncts are summarized in Chapter 2<strong>and</strong> Briggs<br />
(1998). Like special, highly coloured barley malts mash tun adjuncts are deficient or<br />
totally lacking in the enzymes needed to convert the starch or degrade nitrogenous<br />
substances during mashing. Like the unmodified regions of pale malts, <strong>and</strong> especially chit<br />
malts, many adjuncts retain their cellular structure which must be disrupted by cooking or<br />
milling to allow enzymes access to the starch <strong>and</strong> protein during mashing. Malt extracts,<br />
sugars <strong>and</strong> syrups may be used as copper adjuncts to provide more extract, to create highgravity<br />
worts, to `dilute' soluble nitrogen <strong>and</strong> to adjust wort fermentabilities. In general,<br />
adjuncts are used to provide relatively inexpensive extract, to modify beer character<br />
(often resulting in less body, a more `delicate' <strong>and</strong> bl<strong>and</strong> flavour or a less strong<br />
character), <strong>and</strong> to dilute the levels of lipids, polyphenols <strong>and</strong> soluble proteins from the<br />
malt giving a more haze-resistant, <strong>and</strong> sometimes more flavour-stable beer <strong>and</strong> to<br />
improve the head of the beer (Briggs, 1998; Martin, 1978).<br />
The proportions of adjuncts used in mashes vary widely, even within one country. In<br />
Bavaria no adjuncts may be used, in the UK 0 25% of wort extract may be derived from<br />
adjuncts (including copper adjuncts), while in the USA <strong>and</strong> Australia levels of 40 50%<br />
are common <strong>and</strong> sometimes may be higher. It is not possible to switch between adjuncts<br />
at will because the h<strong>and</strong>ling equipment <strong>and</strong> processing conditions that each require are<br />
different. Mash tun adjuncts can cause some brewing problems <strong>and</strong> extra costs, including<br />
the need for extended mashing schedules <strong>and</strong>/or the provision of a cereal cooker, the need<br />
for additions of microbial enzymes, slow wort separations <strong>and</strong> difficulties with beer<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
filtration. As the proportion of adjunct in the mash increases so, at some stage, the<br />
quantities of enzymes available from the malt become inadequate for achieving a good<br />
starch conversion, <strong>and</strong> the recovery of extract declines, the fermentability of the wort<br />
falls, the levels of soluble nitrogen <strong>and</strong> free amino nitrogen, <strong>and</strong> even some inorganic<br />
substances may fall to such a level that yeast growth <strong>and</strong> fermentation may be impaired.<br />
In North America, using the double-mashing system (Section 4.2), up to 60% of maize or<br />
sorghum grits may be used in conjunction with highly enzymic, nitrogen-rich malt. Even<br />
with this system it may be advantageous to use microbial enzymes, such as heat-stable -<br />
amylase, to liquefy the grits in the cooker or a fungal saccharogenic amylase, pullulanase<br />
or amyloglucosidase to adjust wort fermentability.<br />
Green malt, used experimentally in simple infusion mashes, can convert large<br />
proportions of adjuncts exceptionally well giving highly fermentable worts with low<br />
levels of proanthocyanidins that give highly haze-resistant beers (MacWilliam et al.,<br />
1963; Briggs et al., 1981). However, the difficulties of h<strong>and</strong>ling green malt, coupled with<br />
`raw-grain' flavour, have prevented this being used. In <strong>practice</strong>, highly enzymatic, highnitrogen<br />
pale malts made from selected varieties of barley, are best for converting<br />
adjuncts (Allen, 1987; Halford <strong>and</strong> Blake, 1972).<br />
The adjuncts used in simple infusion mashes either contain starches with low<br />
gelatinization temperatures or have been pre-cooked to pre-gelatinize the starch. These<br />
include raw barley, (sometimes regarded as the `natural adjunct'), wheat flour, torrified<br />
barley or wheat <strong>and</strong> flaked maize or rice grits. In addition, flaked barley or wheat may be<br />
used. Flaked maize <strong>and</strong> flaked rice were very popular with British brewers, but they are<br />
little used now, because of their costs. Some pre-cooked cereals give lower extracts than<br />
expected because during cooking some of the amylose is induced to crystallize <strong>and</strong> so<br />
become enzyme-resistant (Home et al., 1994). Provided that adjuncts are not used in<br />
excessive amounts, they are well mixed in with the malt at mashing in <strong>and</strong> the malt<br />
contains an adequate level of enzymes, they are relatively easy to use. The torrified grains<br />
can be milled with the malt. However, raw barley is tough <strong>and</strong> may need to be wet-milled<br />
or ground separately in a specially adjusted mill. `Barley brewing' is considered later.<br />
Wheat flour, like raw barley, can slow wort separation to a serious extent <strong>and</strong> although it<br />
has been used at malt replacement levels of up to 36% (Briggs et al., 1981; Geiger, 1972),<br />
in conventional infusion mashing levels of 5 10% are more usual. The inconvenience of<br />
the slow wort separation is partly offset by enhanced head formation <strong>and</strong> beer stability<br />
Wort filtration problems are reduced by grinding the wheat more coarsely, (e.g by<br />
hammer-milling it with a 2 3 mm screen) <strong>and</strong> by adding a mixture of pentosanase <strong>and</strong><br />
cellulase enzymes to the mash (Forrest et al., 1985).<br />
Highly purified wheat starch is not troublesome but the technical grade usually used has<br />
associated protein <strong>and</strong> other materials which slow run-off. Higher extracts are recovered if<br />
wheat starches or flours are pre-soaked or pre-cooked at about 85 ëC (185 ëF) to allow<br />
gelatinization <strong>and</strong> liquefaction. The material is not boiled to avoid frothing. The flours are<br />
mashed-in at cool temperatures to avoid clumping. Raw barley, flaked barley <strong>and</strong>, to a<br />
lesser extent, torrified barley, like under-modified barley malts, release -glucan into the<br />
wort. This is associated with (but is not necessarily the major cause of) slow wort<br />
separation, but it causes worts to be viscous, it slows beer filtration, <strong>and</strong> sometimes causes<br />
hazes <strong>and</strong> the separation of polysaccharide gels from strong beers. To minimize these<br />
problems it is necessary to use malt containing adequate levels of -glucanase, or to add<br />
microbial -glucanase or cellulase to the mash (Crabb <strong>and</strong> Bathgate, 1973). In temperatureprogrammed<br />
mashing an appropriate low-temperature rest is used to allow the malt enzyme<br />
to act, or a heat-stable microbial -glucanase or cellulase can be added to the mash.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
In double-mashing the adjuncts used are usually maize-, or rice-grits, although sorghum<br />
grits, cleaned sorghum grains <strong>and</strong> other cereal preparations can be used. Rice grits tend to<br />
give a`drier' character to abeer, while maize grits confer a`rounder' <strong>and</strong> more `full'<br />
impression. Cooking is essential when the adjuncts have starches with high gelatinization<br />
temperatures(Table2.3).Ricegrits<strong>and</strong>floursareveryvariableintheirqualities<strong>and</strong>cooked<br />
viscositiesmayvary100-fold<strong>and</strong>run-offtimesthree-fold(Tengetal.,1983).Ricegritsare,<br />
perhaps, the most difficult adjuncts to use. On cooking, arice-mash may set to agel <strong>and</strong><br />
becomeunmanageable.Sometimesitmaybepossibletobuyriceofanamedvariety,having<br />
favourablebrewingcharacteristics,butoftenthisisnotpossible,asonlymixturesofricesof<br />
unnamed varieties are available. Extract recovery is inversely proportional to the gel point,<br />
so ameasurement of the latter can be used to give an indication of the brewing value of a<br />
batch <strong>and</strong> how it should be mashed (Teng et al., 1983). If barley malt is the source of the<br />
liquefying -amylase then at temperatures over 78ëC (172.4ëF) the enzyme is quickly<br />
inactivated while some of the starch will not have been gelatinized. It may be necessary to<br />
employcomplexmashingscheduleswithheating,cooling<strong>and</strong>decoctions<strong>and</strong>morethanone<br />
addition of malt (Kunze, 1996; Narziss, 1992a). Where the use of microbial enzymes is<br />
permittedtherice-mashmaybesupplementedwithacalciumsalt(e.g.100mgCa/litre)<strong>and</strong><br />
the pH adjusted to 6.0. Then aheat-stable bacterial -amylase is added <strong>and</strong> the mixture is<br />
heated, with arest at the best temperature for the enzyme (e.g. 85ëC; 185ëF) to boiling. It<br />
may be preferable, <strong>and</strong> more economic, to use two enzyme preparations, aless expensive<br />
preparation of -amylase with acomparatively low temperature optimum (75ëC; 167ëF)<br />
<strong>and</strong> athermostable enzyme with an optimum temperature of 90ëC (194ëF), <strong>and</strong> then carry<br />
out the heating with rests at the two temperature optima. Cooking may be carried out under<br />
pressure, but at the elevated temperatures attained unwanted flavours, involving the<br />
formation of sulphur compounds, may be formed. At the end of the cooking period the pH<br />
should be adjusted to about 5.5, before mixing with the malt mash.<br />
Sorghum <strong>and</strong> maize grits are used as adjuncts with barley malt mashes. In some areas<br />
of Africa there is pressure to use indigenous cereals to make beers <strong>and</strong> the use of<br />
imported barley malt has been restricted or prevented. Some clear, lager-style beers have<br />
been made using malted sorghum. The poor <strong>and</strong> irregular quality of this material <strong>and</strong> the<br />
high gelatinization temperature of its starch combined with adeficiency of desirable<br />
enzymes means that microbial enzymes are routinely used in making most or all of the<br />
commercially prepared beers (Demuyakar et al., 1994; Hallgren, 1995; Lisbjerg <strong>and</strong><br />
Nielsen, 1991; Little, 1994; MacFadden, 1989; Muts, 1991; Chapter 16). Since added<br />
enzymes are needed it is economic to use little or no sorghum malt <strong>and</strong> to mash<br />
exclusively with whole sorghum grains or sorghum grits <strong>and</strong>/or maize grits. With these<br />
mashes it is necessary to use mash filters for wort separation. The composition of<br />
sorghum grains is very variable, but may yield extracts of about 70%. Grits can yield an<br />
extract of over 90% <strong>and</strong>, at 1%, their lipid content is less than that of the grains at about<br />
3.5%. It is best to use pale grains. If the tannin-rich, coloured, `birdproof' types must be<br />
used then they should be given an alkaline wash with calcium hydroxide, sodium<br />
hydroxide or ammonia solutions to extract as much tannin as possible or a formaldehyde<br />
wash may be used to bind the tannins, although this is now disliked. Failure to control the<br />
tannins results in the inhibition of enzymes during mashing <strong>and</strong> possibly flavour<br />
problems.<br />
Sometimes a proportion of sorghum malt is included in the mashes to enhance the<br />
levels of assimilable nitrogen in the wort (Bajomo <strong>and</strong> Young, 1993). When making an<br />
all-grits mash the problem is to obtain the maximum extract, having the correct degree of<br />
fermentabiliy, with an adequate level of FAN to support yeast growth <strong>and</strong> fermentation.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
This is achieved by a careful choice of enzymes <strong>and</strong> temperature programmes. For<br />
example, the grits suspension is adjusted to pH6, <strong>and</strong> a calcium salt is added at 50 ëC<br />
(122 ëF; Lisbjerg <strong>and</strong> Nielsen, 1991), followed by an addition of a thermostable bacterial<br />
-amylase. The mash is heated to 76 ëC (168.8 ëF) then, after a 15 min. rest, is heated to<br />
boiling <strong>and</strong> is boiled for 30 min. It is then cooled to 52 ëC (125.6 ëF) by adding cold water.<br />
Then preparations of a protease <strong>and</strong> a saccharogenic fungal amylase are added. Following<br />
a 60 min. rest the pH is adjusted to 5.5, then the temperature is raised to 60 ëC (140 ëF)<br />
<strong>and</strong> this is held for 60 min. Finally, after increasing the temperature to 78 ëC (172.4 ëF)<br />
<strong>and</strong> a 10 min. hold, the mash is filtered. Such worts, if no sorghum malt has been used,<br />
are deficient in assimilable nitrogen <strong>and</strong> minerals <strong>and</strong> so need to be supplemented with<br />
`yeast foods' or autolysed yeast <strong>and</strong> perhaps mineral salts, including zinc, to achieve<br />
good fermentations.<br />
In the 1970s efforts were made to use increasing amounts of raw barley as an adjunct.<br />
Moderate amounts of various barley adjuncts are still used, but the use of 70%, or more,<br />
raw grain mashes (with attempts to use 100%), has been discontinued. The problems<br />
encountered in developing `barley brewing' can be summarized as difficulties in milling,<br />
inadequate extract recoveries, poorly fermentable worts, insufficient levels of assimilable<br />
nitrogen, poor foaming characteristics of the final beers, problems due to high levels of -<br />
glucans (slow wort separation, viscous worts, slow beer filtration, hazes <strong>and</strong> gel<br />
formation), higher wort pH values, lack of the correct beer characters <strong>and</strong> colours <strong>and</strong> the<br />
need for microbial enzymes <strong>and</strong> excessively long mashing times. On the other h<strong>and</strong> hop<br />
utilization increased when the worts were boiled <strong>and</strong> the beers were more resistant to<br />
haze formation. The barley had to be carefully screened <strong>and</strong> cleaned (<strong>and</strong> sometimes<br />
washed), <strong>and</strong> preferably had a low nitrogen content. Some varieties were preferable to<br />
others but inexpensive feed-grade grain might be used. In general the use of up to 50%<br />
barley in grists was regarded as successful, but the use of larger proportions seemed to<br />
increase the problems unduly. Wet milling was widely employed but dry milling of<br />
moisture conditioned grain <strong>and</strong> other methods were used (Wieg, 1987; Allen, 1987;<br />
Button <strong>and</strong> Palmer, 1974; Brenner, 1972).<br />
Various enzyme mixtures were employed. The -amylase, -glucanase <strong>and</strong> protease<br />
mixture from Bacillus subtilis was widely used. Wort fermentabilities were adjusted with<br />
additions of fungal saccharogenic amylase, pullulanase or amyloglucosidase or the use of<br />
higher proportions of pale, enzymic malts. Since that time other useful enzymes have<br />
become available. A successful mashing schedule, using 70% raw barley <strong>and</strong> additions of<br />
bacterial enzymes, involved wet milling <strong>and</strong> adding enzymes at mashing-in, which was at<br />
room temperature. Then the temperature was increased to 50 ëC (122 ëF). This was held<br />
for 60 min. Further increases were followed by rests at 63 ëC (145.4 ëF)/75 min., 65 ëC<br />
(149 ëF)/40 min., 68 ëC (154.4 ëF)/20 min. <strong>and</strong> then, after heating to 75 ëC (167 ëF)<br />
transferring to the lauter tun (Button <strong>and</strong> Palmer, 1974). The duration of this process was<br />
about five hours, which is too long to be economic for many breweries.<br />
4.3.4 The influences of mashing temperatures <strong>and</strong> times on wort quality<br />
A knowledge of the influences of mashing temperatures <strong>and</strong> times is essential to allow<br />
the logical choice of mashing conditions. The results of studies, made over many years,<br />
do not agree exactly. This is to be expected since different malts, grinds <strong>and</strong> thicknesses<br />
of the mashes were used. This section concentrates on all-malt mashes. Relative to<br />
laboratory mashes, brewery mashes are more concentrated (have a lower liquid/grist<br />
ratio), they are made with different degrees of milling, with salts in the brewing liquor<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>and</strong> sometimes with pH adjustments. Under brewhouse conditions many key enzymes<br />
remain active for much longer than expected from laboratory studies. Some enzyme<br />
inhibition can occur through product inhibition <strong>and</strong> alimiting supply of `free' water (see<br />
next section). The simplest mashes are isothermal, <strong>and</strong> these are considered first. In<br />
production mashing, temperatures are often adjusted to allow temperature-sensitive<br />
enzymes to act before the temperature is increased to destructive levels. Temperature<br />
adjustments may be achieved by direct steam injection, underletting, decoctions, direct<br />
heating through the walls of the vessel, or the addition of hot water. At the end of the<br />
process, during wort recovery, the temperature is increased by the hot sparge liquor.<br />
Increasing the mash temperature increases the rate of chemical <strong>and</strong> enzyme catalysed<br />
reactions, acceleratestheratesofdenaturation<strong>and</strong> precipitationofproteins(including the<br />
inactivation of enzymes), accelerates dissolution <strong>and</strong> diffusion processes, accelerates<br />
mixing <strong>and</strong>, at least above acertain temperature, causes the gelatinization of starches <strong>and</strong><br />
(at least during decoctions <strong>and</strong> adjunct boiling) disrupts the cellular structure of<br />
unmodified cereal endosperm tissues. Mixtures of enzymes are active in mashing <strong>and</strong><br />
thesehavearangeofwidelydifferenttemperaturesensitivities.Enzymeactivitiesdecline<br />
as mashing proceeds, <strong>and</strong> so the temperature `optima' occur at lower temperatures as<br />
mashing proceeds (Fig. 4.1). Figures 4.7 <strong>and</strong> 4.8 illustrate that the optimum temperature<br />
for the production of permanently soluble nitrogen, which is dependent on amixture of<br />
enzymes, alters with the pH. Some of the substances extracted during mashing are<br />
preformedinthemalt.Thesemayormaynotbealteredasmashingproceeds<strong>and</strong>theyare<br />
joinedbymaterialssolubilizedbyenzyme-catalysedhydrolyticreactions.Someestimates<br />
of temperature optima are shown in Table 4.4.<br />
Permanently soluble nitrogen (mg/100 g malt)<br />
300<br />
250<br />
200<br />
150<br />
100<br />
50<br />
3 h<br />
2 h<br />
1 h<br />
0.5 h<br />
0.25 h<br />
Mashing temperature (°F)<br />
86 104 122 140 158 176<br />
0<br />
0 10 20 30 40 50 60 70 80<br />
Mashing temperature (°C)<br />
Fig. 4.7 The influence of mashing time on the temperature optimum of the production of<br />
permanently soluble nitrogen (data of Windisch <strong>and</strong> Kolbach; after Moll et al., 1974). With<br />
increasing incubation time the temperature `optimum' declines (compare Fig. 4.1).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Mashing time
Fig. 4.8 The influence of mashing temperature on the pH optimum of permanently soluble<br />
nitrogen formation (data of Windisch <strong>and</strong> Kolbach; after Moll et al., 1974).<br />
Table 4.4 Temperature optima for some mash processes, carried out for 2±3 h. Data from various<br />
sources (Briggs et al., 1981; Hind, 1950; Hopkins <strong>and</strong> Krause, 1947). The values can be<br />
substantially different under different mashing conditions<br />
Process ëC ëF<br />
Highest extract (mainly starch conversion) 65±68 149±154.4<br />
Fastest saccharification (dextrinization) 70 158<br />
Highest yield of reducing sugars 60±63 140±145.4<br />
Highest yield of fermentable extract 65 149<br />
Highest percentage fermentability (%) 63 145.4<br />
Highest yield of permanently soluble nitrogen<br />
(Mash times 1±3 h. Higher temp. optima for<br />
more concentrated mashes) 50±55 (60) 122±131 (140)<br />
Highest yield of formol-nitrogen 50±55 122±131<br />
Highest PSN minus formol-N 55±60 131±140<br />
Highest yield of `acid buffers' 45±55 113±131<br />
Maximum activity of -amylase 70 158<br />
Maximum activity of -amylase 60 140<br />
Maximum activity of -glucanase 40 104<br />
Maximum activity of phytase 50±60 122±140<br />
As noted previously, these optima are not true `constants'. At mashing temperatures<br />
the survival <strong>and</strong> activity of the proteolytic system of enzymes is extremely dependent on<br />
mash thickness. -Amylase is less stable <strong>and</strong> both -amylase <strong>and</strong> limit dextrinase are<br />
more stable than might be predicted from their behaviours in pure solutions. Most of the<br />
extract formed during mashing comes from the conversion of starch to a mixture of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Extract (%)<br />
70<br />
50<br />
30<br />
10<br />
95 80 70 60 50<br />
Temperature (°C)<br />
soluble sugars, oligosaccharides <strong>and</strong> dextrins. It is not true that starch can be converted<br />
only at or above its gelatinization temperature, although this occurs most rapidly after the<br />
granules have been disrupted. It has been known for many years that in mashes made at<br />
55ëC (131ëF) over 90% of the potential extract can be recovered in two hours, although<br />
this temperature is well below the gelatinization temperature of barley starch (Table 2.3).<br />
Even at lower temperatures some starch conversion occurs, so malt enzymes can slowly<br />
degrade malt-starch granules (Fig. 4.9). When the time-courses of extract <strong>and</strong><br />
permanently soluble nitrogen formation are followed it is seen that in amash made at<br />
65.5ëC (150ëF) the extract rises rapidly at first <strong>and</strong> most of the extract has been<br />
recovered in one hour but, in this instance, the maximum is not recovered until after<br />
1.5 2h. (Table 4.5). This limit is set by the nearly complete solubilization of the starch.<br />
In contrast the extract of the 49ëC (120.2ëF) mash is still rising after three hours, <strong>and</strong><br />
extract recovery is not nearly complete. The amount of nitrogen solubilized during<br />
mashing never approaches all the nitrogenous substances in the malt, so the halt in the<br />
riseinPSNinthe65.5ëCmashafterabouttwohoursisduetoenzymeinactivation.Inthe<br />
49ëC mash the PSN is still increasing after three hours, <strong>and</strong> the value exceeds that<br />
achieved at 65.5ëC, provingthatadepletionofinitially insolublenitrogenous materials is<br />
not limiting. Even in three hours at 49ëC less that 40% of the total nitrogen has been<br />
40<br />
80<br />
Time (min.)<br />
Fig. 4.9 The interrelationships between the yields of extract, the mashing duration <strong>and</strong> the<br />
mashing temperature (after Schur et al., 1975).<br />
Table 4.5 Changes in yields with time of extract <strong>and</strong> permanently soluble nitrogen (PSN) in<br />
mashes made at two different temperatures (data of H. T. Brown, via Hind, 1950)<br />
Time (min.) 15 30 60 90 120 180<br />
Mash at 49 ëC (120 ëF)<br />
Extract (%) 20.3 24.8 28.2 30.3 34.0 37.1<br />
PSN (% TN) 24.6 28.0 32.2 34.6 36.5 39.0<br />
Mash at 65.5 ëC (150 ëF)<br />
Extract (%) 63.4 67.1 69.4 69.7 69.7 69.0<br />
PSN (% TN) 27.7 30.7 32.9 33.7 34.6 34.6<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
160
Table 4.6 Extracts <strong>and</strong> fermentable extracts obtained in isothermal mashes during different<br />
incubation periods. (Data of Windisch, Kolbach und Schild, via Hopkins <strong>and</strong> Krause, 1947)<br />
Mashing period (min.) 15 30 60 120 180<br />
60 ëC (140 ëF)<br />
Extract (%) 50.2 53.4 57.2 60.7 62.2<br />
Fermentable extract (%) 36.0 39.0 43.1 47.9 50.2<br />
65 ëC (149 ëF)<br />
Extract (%) 60.6 62.2 62.8 63.6 63.6<br />
Fermentable extract (%) 44.2 46.6 48.5 50.7 51.7<br />
70 ëC (158 ëF)<br />
Extract (%) 61.2 62.5 62.9 63.4 63.6<br />
Fermentable extract (%) 40.9 42.0 41.6 42.2 42.7<br />
brought into solution. However, extract recoveries can be nearly complete at<br />
temperatures as high as 80ëC (176ëF), but fermentability may be as low as 30%,<br />
compared to 70%, or more, in worts from mashes made at 65ëC (149ëF) (Fig. 4.13 on<br />
page 117; Muller, 1991).<br />
At 80ëC (176ëF) sufficient -amylase remains for starch liquefaction <strong>and</strong> partial<br />
dextrinization to occur while -amylase <strong>and</strong> other heat labile enzymes are so rapidly<br />
destroyed that little saccharification can take place. In Table 4.6 the maximum extract<br />
recovery was nearly achieved in three hours, mashing at 60ëC (140ëF) <strong>and</strong> was achieved<br />
in about two <strong>and</strong> ahalf hours at 65 <strong>and</strong> 70ëC (149 <strong>and</strong> 158ëF). On the other h<strong>and</strong> the<br />
fermentable extract was still rising after three hours in the lowest temperature mash, was<br />
rising more slowly in the 65ëC (149ëF) after three hours, but had stopped increasing<br />
between one <strong>and</strong> two hours in the mash made at the highest temperature. Thus mash<br />
temperatures must be very carefully controlled if worts of one quality are to be produced.<br />
When the extract yields of isothermal mashes, made at temperatures between 0ëC (32ëF)<br />
<strong>and</strong> 80ëC (176ëF), are compared it is seen that the extracts increase relatively little with<br />
increasing temperature until 45 50ëC (113 122ëF) is reached. Then extracts rise<br />
sharply with increasing temperature up to about 55 60ëC (131 140ëF), then the rate of<br />
increase falls sharply <strong>and</strong> the maximum is achieved at 62 66ëC (143.6 150.8ëF). At<br />
higher temperatures there is aslow decline in extract recovery, at least up to 80ëC<br />
(176ëF;Windischetal.,1932). Therelationshipbetweenextract yield<strong>and</strong>amorenarrow<br />
temperature range for different times is shown in Fig. 4.9. By using different series of<br />
increasing temperatures, in temperature-programmed mashes, wort composition can be<br />
adjusted to auseful extent, for example, to produce beers with different alcohol contents.<br />
Mashing for extended periods at low temperatures favours the formation of soluble<br />
nitrogen, the optimum temperature depending on the time (Fig. 4.7). Carbohydrates are<br />
the main contributors to extract. The formation of the major groups of carbohydrates<br />
during atemperature-programmed mash are illustrated in Fig. 4.10 (Enevoldsen, 1974).<br />
Notice that starch hydrolysis must have begun before the temperature reached 63 ëC<br />
(149 ëF). The `rest' temperatures used in mashing are chosen with reference to the<br />
temperature optima of key groups of enzyme-catalysed reactions. The low-temperature<br />
rests at about 45 50 ëC (113 122 ëF) are needed with undermodified malts when the<br />
breakdown of proteins <strong>and</strong> -glucans is to be encouraged. The rests at about 65 ëC<br />
(149 ëF) are to maximize starch conversion <strong>and</strong> production of fermentable sugars. The<br />
fermentable group of sugars includes glucose (4.1), fructose (4.2), sucrose (4.3) <strong>and</strong><br />
maltotriose (4.5), but the major component is maltose (4.4). The non-fermentable<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Extract (g/100 ml)<br />
15<br />
10<br />
5<br />
52°C<br />
(125.6°F)<br />
63°C<br />
(145.4°F)<br />
0<br />
0 30 50 110 160<br />
Fig. 4.10<br />
Mashing time (min.)<br />
Changes in the yield of total extract (TE), the yield of total carbohydrate (TC), the<br />
fermentable sugars (FS) <strong>and</strong> the unfermentable dextrins (D) during atemperature-programmed<br />
mash (after Enevoldsen, 1974).<br />
carbohydrate fraction is chiefly a complex mixture of dextrins. Mannose (4.8) <strong>and</strong><br />
galactose (4.9) are not released into solution. The extract increases at 52ëC (125.6ëF) but<br />
increases much more rapidly as the temperature increases to 63ëC (145.4ëF) <strong>and</strong> is held<br />
at this temperature. It continues to increase slightly up to the final temperature of 78ëC<br />
(172.4ëF) (Fig. 4.10).<br />
The changes that occur in yields of soluble nitrogen during temperature-programmed<br />
mashing are illustrated in Fig. 4.11. In the mash made at 35ëC (95ëF) the highest level of<br />
TSN was achieved after about 100min., then as the temperature continued to increase the<br />
level fell, due to the denaturation <strong>and</strong> precipitation of solubilized proteins. Beginning the<br />
programme at ahigher temperature, 50ëC (122ëF) yielded the same final amount of TSN<br />
at the end of the mash, but mashing in at 65ëC (149ëF) reduced the final amount of TSN<br />
significantly. The amounts of amino nitrogen formed fell as the mashing in temperature<br />
increased, so the mixture of nitrogenous substances in solution was altered by changes in<br />
the mashing regime. This result is due to the different temperature sensitivities of the<br />
enzymes involved. Prolonged mashing regimes are usually needed when under-modified<br />
malts are used, <strong>and</strong> the advantage in terms of yield of TSN is also shown in Fig. 4.11.<br />
The production of amino acids (measured by two methods) during a different<br />
temperature-programmed mash is shown in Fig. 4.12. It was long believed that<br />
proteolysis in mashing ceased at temperatures above about 60ëC (140ëF). This is not so.<br />
Proteolysis is most rapid at lower, `conventional' protein rest temperatures, but it does<br />
not cease immediately the temperature rises. At these lower temperatures phytase <strong>and</strong> -<br />
glucanase also continue to act. By mashing amalt in different ways worts with different<br />
qualities are produced (Table 4.7). By encouraging proteolysis the colours of worts are<br />
increased, probably because the elevated levels of nitrogen-containing substances favour<br />
melanoidin formation during the copper boil. More complete proteolysis reduces the<br />
amount of break formed <strong>and</strong> this enhances the hop utilization, since less of the bitter<br />
substances are deposited with the coagulated protein (trub). Less trub (sludge) is formed<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
T.E.<br />
T.C.<br />
D<br />
F.S.<br />
78°C<br />
(172.4°F)
TSN (mg/100 g of dry malt)<br />
α-Amino N (mg/100 g of dry malt)<br />
850<br />
800<br />
750<br />
700<br />
650<br />
600<br />
550<br />
500<br />
200<br />
190<br />
180<br />
170<br />
160<br />
150<br />
(a)<br />
(b)<br />
35°C<br />
95°F<br />
35°C Undermodified malt<br />
0 60<br />
Time (min.)<br />
120 180<br />
50°C<br />
122°F<br />
65°C<br />
149°F<br />
158°F<br />
75°C<br />
167°F<br />
0 60<br />
Time (min.)<br />
120 180<br />
Fig. 4.11 Increases in the levels of the (a) total soluble nitrogen <strong>and</strong> (b) the -amino nitrogen<br />
during temperature-programmed mashing. The different mashes were started at the temperatures<br />
shown <strong>and</strong> then the programme was completed (after Narziss, 1977). The discontinuous line in (a)<br />
shows the results obtained with a different, undermodified malt.<br />
in worts from decoction mashes, probably because some of the protein has been<br />
denatured <strong>and</strong> precipitated in the boiled parts of the goods.<br />
4.3.5 Non-malt enzymes in mashing<br />
The mainly microbial, non-malt enzymes used in brewing are indicated in Chapter 2. In<br />
this section only the addition of enzymes to mashes is considered. Because of the absence<br />
of agreed methods of assay <strong>and</strong> the presence of `extra' unspecified <strong>and</strong>/or unquantified<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
70°C<br />
35°C<br />
50°C<br />
65°C<br />
35°C<br />
50°C<br />
65°C
Amino acids (μmoles/g dry wt)<br />
80<br />
70<br />
60<br />
50<br />
45°C<br />
(113°F)<br />
55°C<br />
(131°F)<br />
TCA-sol. α-NH 2<br />
0 0 30 60 90 120 150<br />
Mashing time (min.)<br />
70°C<br />
F.A.A.<br />
(158°F)<br />
Fig. 4.12 The formation of amino acids during a temperature-programmed mash (after Enari,<br />
1974). The `amino acids' were estimated in two different ways, which gave discordant results.<br />
Although amino acid liberation proceeded fastest at low temperatures it was still going on at the end<br />
of the mash at the highest temperature. (FAA, free amino acids; TCA sol- -NH 2, amino groups<br />
soluble in trichloroacetic acid.)<br />
enzymes in enzyme preparations it is essential that brewers test each preparation on the<br />
laboratory <strong>and</strong> pilot-plant scales before introducing it into the brewhouse. The dose-rate<br />
is critical for achieving a particular result, <strong>and</strong> this must be financially worthwhile.<br />
Difficulties arise when a preparation that has been correctly st<strong>and</strong>ardized on one enzyme<br />
activity is used but the activity of an important `secondary' enzyme differs between<br />
batches or declines at a different rate during storage. It is possible to store particular<br />
preparations at `room temperature' for some months but in principle, all enzyme<br />
preparations should be stored cold to minimize rates of deterioration. The use of added<br />
enzymes when mashing adjuncts has been considered <strong>and</strong> in the cases of barley- <strong>and</strong><br />
grits-mashing there is a need for enzymes to cause starch liquefaction <strong>and</strong><br />
saccharification, -glucan degradation <strong>and</strong> protein hydrolysis. The raw barley contains<br />
bound <strong>and</strong> soluble -amylase, the bound form of which may, at least experimentally, be<br />
partly activated <strong>and</strong> solubilized by additions of the amino acid cysteine hydrochloride, or<br />
other thiol-containing substances or sulphite, making this enzyme more useful in<br />
mashing. Raw barley contains proteins which can inhibit some microbial proteases. The<br />
difficulty in obtaining sufficient assimilable nitrogen from barley- <strong>and</strong> grits-mashes<br />
underlines the absence of inexpensive peptidase preparations suitable for use in mashing.<br />
Enzymes may be added to mashes to speed starch conversion, to degrade -glucans, or<br />
to accelerate wort run-off. Various enzymes are adequately active at mash pHs but these<br />
often do not coincide with the pH optima of the enzymes. Usually the mash pH is adjusted<br />
only once at the onset of mashing. The exception is when grits are being cooked the pH is<br />
adjusted to 6 to allow a heat-stable bacterial -amylase to act. Afterwards the pH must be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.7 A comparison of worts made in a Bavarian brewery, with an undermodified, 1931 malt (u) <strong>and</strong> a well-modified, 1932 malt (m) using four different<br />
mashing programmes (data of LuÈers et al., 1934). Pi, inorganic phosphate as % of the total in the malt. TSN, total soluble nitrogen, <strong>and</strong> amino-nitrogen as % of dry<br />
extract. In all heating periods, including decoctions, there were rests at 50 ëC (122 ëF), <strong>and</strong> 65±68ëC (149±154.4 ëF). Worts were collected at 76 ëC (168.8 ëF)<br />
Wort Wort<br />
Extract attenuation viscosity TSN Amino-N Pi Sludge<br />
(% dm) (% dm) (%) (relative) (% dry extr.) (% dry extr.) (%, total) (kg dry)<br />
Malt modification u m u m u m u m u m m u m<br />
Two-mash process. Mash in at 50 ëC<br />
(122 ëF). Decoctions to 67 ëC (152.5 ëF)<br />
<strong>and</strong> then 76 ëC (169 ëF) 77.5 78.8 66.2 65.6 1.79 1.80 0.87 0.85 0.29 0.26 83.5 11.5 13.6<br />
Three-mash process. Mash in at 35.5 ëC<br />
(96 ëF). Decoctions to 50 ëC (122 ëF),<br />
65 ëC (149 ëF) <strong>and</strong> 76 ëC (169 ëC). 78.3 79.3 63.4 64.5 1.71 1.79 0.93 0.89 0.31 0.29 86.4 13.6 9.9<br />
`High-quick mash process'. Mash in<br />
at 68 ëC (154.5 ëF), then raise to 76 ëC<br />
(169 ëF). 76.2 78.4 67.4 67.5 1.76 1.79 0.92 0.78 0.25 0.21 79.4 26.7 17.5<br />
Temperature-programmed, infusion-mash.<br />
Mashed in at 35 ëC (95 ëF). Heat with<br />
steam-coils, with rests at 50 ëC (122 ëF),<br />
67 ëC (152.5 ëF) <strong>and</strong> 76ëC (169 ëF). 78.2 78.8 66.8 66.5 1.68 1.70 0.94 0.86 0.28 0.27 86.4 26.9 22.8<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
eadjusted downwards to about 5.5 before adding to the malt mash or saccharifying with a<br />
fungal saccharogenic amylase or -amylase, with or without pullulanase. The low<br />
temperature stabilities of these enzymes is inconvenient <strong>and</strong> may necessitate cooling the<br />
mash to about 50 55ëC (122 131ëF) to allow them to act to an adequate extent. The<br />
choice of which bacterial -amylase to use is largely dictated by their stabilities. The<br />
Bacillussubtilisenzymeworkswellat70ëC(158ëF)<strong>and</strong>themorestableenzymefromthe<br />
B. subtilis var. amyloliquefaciens, is useful in cookers. It is inactivated on boiling, <strong>and</strong> this<br />
is an advantage. If amore stable enzyme, from B. licheniformis, is used it must be under<br />
conditionsthatensureitsinactivationduringthecopperboilorthecompositionofthewort<br />
<strong>and</strong>beerwilldriftwithtimeasdegradationoftheresidualdextrinscontinues.Theaddition<br />
of uneconomically large amounts of amixture of bacterial -amylase <strong>and</strong> -glucanase<br />
(specific for mixed-link -glucans) can increase the laboratory extract of asound malt<br />
obtained with an extended mashing schedule, e.g., from 308 to about 313 lë/kg (Albini et<br />
al., 1987). The major use of these preparations is to offset the deficiencies of undermodified<br />
malts <strong>and</strong> the presence of barley adjuncts. Most preparations also contain<br />
proteases which, while they can usefully increase the level of soluble nitrogen in the wort,<br />
can reduce the head formation <strong>and</strong> stability in the final beer.<br />
In modern breweries with very short `turn-round times' it is essential that wort<br />
separation occurs rapidly. Slow wort separation may be associated with using poor or<br />
inhomogeneous malt. Of the various enzymes tested the preparations of the relatively<br />
heat-stable cellulases, e.g., from Trichoderma viride (50 55ëC; 122 131ëF), Penicilliumfuniculosum(65ëC;149ëF)<strong>and</strong>P.emersonii(80ëC;176ëF)seemtobesuccessfulin<br />
accelerating wort separation. These preparations contain complex mixtures of enzymes<br />
that attack mixed-link glucans <strong>and</strong> holocellulose (polysaccharides insoluble in hot<br />
sodium hydroxide solution) <strong>and</strong> may also have pentosanase activity. These preparations<br />
sometimes improve extract recovery. In mashes containing awheat adjunct it may be<br />
necessary to supplement the mash with a mixture of cellulase <strong>and</strong> pentosanase<br />
preparations to achieve asufficiently fast wort separation.<br />
4.3.6 Mashing liquor <strong>and</strong> mash pH<br />
Mashing liquor must be free of taints, <strong>and</strong> must be potable. In addition it must be free of<br />
many substances <strong>and</strong> organisms which might reduce beer quality. The quality of the<br />
water received must be checked regularly, whatever the source <strong>and</strong>, if necessary, it must<br />
be treated to convert it to the proper quality for the beers to be brewed. For mashing <strong>and</strong><br />
beer dilution liquors oxygen may have been removed <strong>and</strong> dilution liquors may have been<br />
charged with carbon dioxide, with or without nitrogen (Chapter 3). The ratio of<br />
temporary to permanent hardness, the total amount of hardness <strong>and</strong> the amounts of ions<br />
that may influence flavour must be regulated (Narziss, 1992a; MacWilliam, 1975; Taylor,<br />
1981). Interactions between calcium (<strong>and</strong> to a lesser extent magnesium) ions <strong>and</strong> wort<br />
components have an important effect on mash <strong>and</strong> wort pH values. Thus bicarbonate ions<br />
effectively remove hydrogen ions <strong>and</strong> so, indirectly, raise the pH:<br />
HCO3 ‡ H ‡ $ H2CO3 ! CO2 " ‡ H2O<br />
Calcium ions (<strong>and</strong> magnesium ions to a lesser extent) interact with mash components<br />
such as inorganic phosphate, phytic acid <strong>and</strong> less phosphorylated inositol phosphates,<br />
peptides, proteins <strong>and</strong> probably with other substances displacing hydrogen ions into the<br />
mash <strong>and</strong> reducing the pH. For example,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
3Ca 2‡ ‡2HPO4 2 !2H ‡ ‡Ca3…PO4† 2 #<br />
The calcium phosphate tends to precipitate <strong>and</strong> precipitates more rapidly from wort at<br />
higher temperatures. Thus the pH of mashes decline faster during decoction mashing <strong>and</strong><br />
later the pH of the wort declinesfurther during the hop-boil.Of the calcium ions added to<br />
amash 40 60% are retained in the spent grains.<br />
Amajor difficulty follows from the habit of measuring the pH of worts or mashes at<br />
roomtemperature<strong>and</strong>assumingthatthesevaluesapplyathighertemperatures,whenthey<br />
do not (Hopkins <strong>and</strong> Krause, 1947). Weak acids, like water (see Appendix), dissociate<br />
more as the temperature rises <strong>and</strong> so the pH values of their solutions fall, like the pH<br />
values of mashes (Table 4.8). Thus at 65 ëC (149 ëF) the pH of a wort is likely to be about<br />
0.35 pH unit lower than at room temperature <strong>and</strong> 0.45 lower at 78 ëC (172.4 ëF). As the<br />
temperature of a mash changes (decoctions, temperature programming, sparging) so will<br />
the pH. These differences are significant, yet in many reports it is unclear if pH values<br />
have been determined at wort- or mash-temperatures or on cooled samples. Probably the<br />
latter is most usual. The pH optimum of -amylase, determined at room temperature, is<br />
about pH 5.3, but its optimum estimated from mashing experiments is often reported to be<br />
about 5.7. This error is due to the pH having been determined on the mash after it was<br />
cooled, when the pH had risen. Because of this difficulty the pH optima of changes<br />
occurring in mashes are a little uncertain (Table 4.9).<br />
Mashing pale malt in distilled water usually gives a wort with a pH of about 5.8 6.0,<br />
this value being maintained by the buffer substances (including phosphates <strong>and</strong> proteins)<br />
Table 4.8 Changes in the pH values of two mashes, made with distilled water <strong>and</strong> moderately<br />
carbonated water, measured at the temperatures shown (after Hopkins <strong>and</strong> Krause, 1947)<br />
Temperature of measurements pH values of the mashes<br />
( ëC) ( ëF) (Distilled water) (Carbonated water)<br />
18 64.4 5.84 6.03<br />
35 95 5.70 5.90<br />
52 125.6 5.65 5.80<br />
65 149 5.50 5.70<br />
78 172.4 5.40 5.55<br />
Table 4.9 `Optimal' pH values for `normal' isothermal infusion mashes made with pale malts<br />
lasting 1±2 h. at 65.5 ëC (150 ëF). Data from various sources. As far as possible the temperatures<br />
(mash temperature, m , <strong>and</strong> cooled wort, w), at which the pH values were determined are indicated<br />
(see text for a warning)<br />
Characteristic `Optimal' pH<br />
Shortest saccharification (dextrinization) time 5.3 m±5.7 w<br />
Greatest extract obtained 5.2±5.4 m?<br />
Greatest extract from a decoction mash 5.3 m±5.6 m?<br />
The most fermentable wort 5.1±5.3 m?; 5.4±5.6 w?<br />
Mash impossible to filter < 4.7<br />
-Amylase most active (+ Ca 2+ ) 5.3 m±5.7 w<br />
-Amylase most active 5.1±5.3 (4.7?)<br />
Maximum yield of PSN 4.4±4.6 m; 4.9±5.1 w<br />
Maximum yield of formol-N 4.4±4.6 m, 4.9±5.2 w<br />
Maximum protease activity (depends on substrate) 4.3 m; 4.6±5.0 m?<br />
Maximum phytase activity about 5.2 m<br />
Carboxypeptidase activity maximal 4.8±5.7<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
from the grist. Infusion mashes are best carried out at pH 5.2 5.4 (mash temperature),<br />
<strong>and</strong> so will give cooled worts with pH values of about 5.5 5.8. It has been recommended<br />
that decoction mashes should not give worts with pH values less than 5.5. Lowering the<br />
pH too much results in increases in the soluble nitrogenous materials but lengthens the<br />
saccharification time <strong>and</strong> lowers the yield of extract. Lowering the pH to the correct<br />
extent, by additions of calcium salts or other means, speeds the rate of starch degradation,<br />
enhances the activities of other carbohydrases <strong>and</strong> the proteolytic mixture of enzymes so<br />
that TSN <strong>and</strong> FAN values are increased <strong>and</strong> wort colour is reduced. The solubility<br />
characteristics of some proteins are altered, the buffering power of the wort is increased<br />
<strong>and</strong>, at later stages of brewing, hop utilization is decreased. Conversely the increase in pH<br />
caused by mashing with waters rich in bicarbonate ions is generally undesirable (Table<br />
4.10). The pH values of mashes are conveniently lowered by mashing with `permanently<br />
hard' water, either natural or that has been `Burtonized' by additions of calcium sulphate<br />
<strong>and</strong>/or calcium chloride (Tables 4.11, 4.12). Often about 100 mg of calcium is added to<br />
each litre of liquor. However, other means may be adopted such as the direct addition of<br />
sulphuric, phosphoric or lactic acids (where this is permitted) or by the use of lactic acid<br />
malts or acidified worts.<br />
Table 4.10 The effects of water hardness on the pH values of the cold water extracts <strong>and</strong> cooled<br />
worts prepared by decoction mashing (after Hopkins <strong>and</strong> Krause, 1947)<br />
Nature of water used pH of CWE or wort<br />
Distilled water; cold water extract (CWE) 6.2±6.3<br />
Wort, water with temporary hardness (about 15 grains CaCO 3/gal.,<br />
214 mg/litre) 5.89<br />
Wort, distilled water 5.76<br />
Wort, water with permanent hardness (about 4 grains CaSO 4/gal.,<br />
57 mg/litre) 5.65<br />
Table 4.11 The amounts of salts (shown in the anhydrous forms) added to some British brewing<br />
liquors (Comrie, 1967). Larger amounts of calcium salts may be added to offset the presence of<br />
bicarbonate ions. The amounts are varied to allow for alterations in mash thickness <strong>and</strong> to achieve<br />
the desired final flavours. (1 grain/imperial gallon (UK) ˆ 14.21 mg/litre)<br />
Pale ales Mild ales Stouts<br />
Salts grains/gal mg/l grains/gal mg/l grains/gal mg/l<br />
CaSO4 15.7±31.5 223±448 5±10 71±142 none none<br />
CaCl 2 6.9±13.7 98±195 7±14 99±199 5.5±11 78±156<br />
MgSO4 2.5 36 2.5 36 2.5 36<br />
NaCl 2.5±5.0 36±71 5±10 71±142 7±12 99±171<br />
Table 4.12 The effects of calcium ions, added as calcium chloride, on the pH, extract <strong>and</strong> soluble<br />
nitrogen fractions given by mashes made with one malt at 65 ëC (149 ëF; Taylor, 1981)<br />
pH Extract TSN FAN<br />
(mg/litre) (litreë/kg) (ppm) (ppm)<br />
AddedCa 2+<br />
0 5.74 287 904 188<br />
100 5.48 291 973 195<br />
200 5.39 292 983 207<br />
300 5.28 292 1062 220<br />
The pH values were measured at room temperature. TSN <strong>and</strong> FAN were adjusted to a wort concentration of SG<br />
1040.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table4.13 Someoftheinfluencesofaddinggypsum(calciumsulphate,CaSO4.2H2O. This contains<br />
23.28% Ca, by weight) to the liquor when mashing malt (data of Hind, via Briggs et al., 1981)<br />
Gypsum added Extract (l ë/kg) Unboiled wort Boiled wort<br />
(mg/litre liquor)<br />
(Apparent) (Corrected Ash Phosphates Phosphate<br />
for ash (mg/100 ml) as P2O5 as P2O5<br />
content) (mg/100 ml) (mg/100 ml)<br />
0 296.7 286.4 138 70 68<br />
380 300.5 289.3 148 63 59<br />
760 302.9 290.4 167 56 54<br />
1140 305.7 290.6 197 54 50<br />
Additions of calcium ions to the mash reduce the quantities of phosphates in solution<br />
but apparently not to undesirable extents (Table 4.13). In addition to the advantages<br />
achieved by favourable pH adjustments the calcium ions stabilize -amylase during<br />
mashing, accelerate wort separation <strong>and</strong> run-off from the mash, assist in break formation<br />
in the hop-boil <strong>and</strong> the beer clarifies better, yeast flocculation is favoured <strong>and</strong> calcium<br />
oxalate crystals (which can be deposited on the walls of fermenters as `beer stone') are<br />
precipitated <strong>and</strong> so the potentially toxic oxalic acid (4.151) does not go forward. In the<br />
beer the calcium oxalate may give rise to haze or initiate gushing. Where the<br />
Reinheitsgebot <strong>and</strong> similar laws operate, the addition of `chemicals' is not permitted <strong>and</strong><br />
with some beers (e.g. Pilsen-style lagers) the brewing liquor must be soft. In these cases<br />
the adjustment of mash pH values is achieved by the use of biologically prepared lactic<br />
acid introduced into the mash either as acid malt or as acidified wort. Acid malts are<br />
prepared in various ways (Briggs, 1998), <strong>and</strong> carry 1 5% (typically 2%) lactic acid <strong>and</strong>,<br />
on being mashed alone, give awort with apH in the range 3.8 4.4. The pKa of lactic<br />
acid is 3.86 at 25ëC. Usual additions are about 5% of the grist but larger quantities may<br />
be used. Some beers derive part of their character from the lactic acid they contain. An<br />
alternative is to acidify unhopped first wort by incubating it with thermophilic lactic acid<br />
bacteria (Lactobacillus delbruÈckii, L. amylolyticus) at 45 47ëC (113 116.6ëF) for<br />
8 71h(Oliver-Daumen, 1988). Batch, semi-continuous <strong>and</strong> continuous acidification<br />
plants are available, the batch type being the most common. It is possible to calculate the<br />
amount of acid needing to be added to achieve adesired reduction in pH.<br />
During lautering (wort collection) buffers are washed out of the mash <strong>and</strong> there is a<br />
tendency for the pH to rise, particularly if bicarbonate is present in the hot (e.g.<br />
75 80ëC; 167 176ëF) sparge liquor. This is highly undesirable as at the higher pH<br />
unwanted polyphenols <strong>and</strong> flavour substances are leached from the goods. This may<br />
make it necessary to treat the weaker last runnings with active charcoal (to remove<br />
unwantedsubstances), beforethey areadded backtoasubsequentmashoraretransferred<br />
tothecoppertobeboiled.ItispreferabletomakesurethatanyriseinpHduringsparging<br />
is minimal by excluding the use of water containing bicarbonates <strong>and</strong> by ensuring that<br />
adequate levels of calcium ions are present.<br />
4.3.7 Mash thickness, extract yield <strong>and</strong> wort quality<br />
Changes in mash thickness (liquor/grist ratio) have significant effects on mash<br />
performance (Hind, 1950; Hopkins <strong>and</strong> Krause, 1947; Harris <strong>and</strong> MacWilliam, 1961;<br />
Muller, 1989; 1991; Table 4.14). Very concentrated mashes, (liquor/grist
Table 4.14 The influence of mash concentration on worts from mashes made at 60 ëC (140 ëF),<br />
with a duration of 180 min. (Data of Windisch, Kolbach <strong>and</strong> Schild, via Hopkins <strong>and</strong> Krause, 1947)<br />
Concentration of mash (water:malt) 2:1 2.7:1 4.0:1 5.3:1<br />
Extract (% dry malt) 71.7 77.0 80.0 79.9<br />
Fermentable extract (% dry malt) 52.3 56.3 58.5 57.8<br />
Fermentable extract (% total extract) 72.9 73.1 73.1 72.3<br />
Permanently soluble nitrogen (% dry malt) 0.57 0.56 0.54 0.53<br />
Formol nitrogen (% dry malt) 0.22 0.21 0.20 0.19<br />
down, worts are more concentrated <strong>and</strong> viscous, TSN <strong>and</strong> FAN are increased <strong>and</strong> more<br />
high molecular weight nitrogenous substances remain in solution, but a lower proportion<br />
of hydrophobic peptides (relative to the amount of extract) are present, causing `high<br />
gravity' beers to have poor head retentions (Bryce et al., 1997). In the concentrated<br />
mashes both the enzymes <strong>and</strong> their substrates are more concentrated. Some enzymes<br />
(proteolytic enzymes, disaccharidases) are more stable in concentrated mashes producing<br />
higher proportions of TSN <strong>and</strong> hexose sugars respectively. At high mashing temperatures<br />
thicker mashes give worts with higher fermentabilities (Muller, 1991; Fig. 4.13). On the<br />
Starch (%) Fermentability (%) Carbohydrate (%)<br />
12<br />
10<br />
8<br />
6<br />
80<br />
60<br />
40<br />
20<br />
0<br />
1.6<br />
1.2<br />
70°, 75°,<br />
80°, 85 °C<br />
70°C<br />
75°C<br />
80°C<br />
85°C<br />
85°C<br />
0.8<br />
80°C<br />
0.4<br />
75°C<br />
70°C<br />
0<br />
50 100 150 200 250 300 350 400<br />
Water (ml)/malt (50 g)<br />
Fig. 4.13 The influences of mash thickness <strong>and</strong> mash temperature, during 1 h. isothermal<br />
mashing, on (upper) the yield of carbohydrate in the wort, (middle) the fermentability of the wort<br />
<strong>and</strong> (lower) the starch present in the wort (after Muller, 1991).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Extract (L°/kg)<br />
300<br />
250<br />
200<br />
150<br />
130<br />
0 1 2 3 4 5 6 7<br />
Water (ml)/Grist (g) ratio<br />
Potato starch<br />
Wheat starch<br />
Controls (all malt)<br />
Barley<br />
Maize starch<br />
Fig. 4.14 The effects of mash thickness on the extract recoveries from mashes made with all-malt<br />
or 50:50 mixtures of malt <strong>and</strong> the adjuncts indicated (data of Muller, 1991).<br />
other h<strong>and</strong>, at `normal' mashing temperatures weaker mashes give more fermentable<br />
worts. The high concentrations of sugars <strong>and</strong> dextrins present in thick mashes can inhibit<br />
the amylases. Enzyme inhibition is due to the reduced availability of free water as well as<br />
to the sugars acting as competitive inhibitors. Brewery worts contain 0 40% more<br />
soluble nitrogen than laboratory analytical worts. It was reported that mashes made with<br />
39% solids give worts with maximum extract yields while worts with the highest<br />
fermentabilities are given by mashes made with 16 32% solids. The effects of mash<br />
concentration on extract yield are also present when adjuncts are included in the mash<br />
(Harris <strong>and</strong> MacWilliam, 1961; Muller, 1991; Fig. 4.14).<br />
As the grist hydrates water is bound, <strong>and</strong> there is a rise in temperature caused by the<br />
release of heat (the `heat of hydration'). As the mash proceeds water is utilized in<br />
hydrolyses, a water molecule being consumed when any bond is split. Some water is<br />
more or less firmly bound (by hydrogen bonding) to starch, to sugars in solution, to -<br />
glucans, to pentosans <strong>and</strong> to other substances reducing the concentration of `free' water.<br />
In all-malt mashes <strong>and</strong> mashes made with 50 : 50 malt <strong>and</strong> barley or wheat starch the<br />
extract recovered falls very sharply as the liquor/grist ratio is reduced below about 2.5<br />
(Fig. 4.14). Generally, altering the liquor/grist ratio at values over 3 has comparatively<br />
minor effects, but these are not necessarily negligible. In a particular case mashing with a<br />
liquor/grist ratio of 2.5 : 1 gave an extract of 291 l ë/kg, while at a ratio of 7 : 1 the extract<br />
was 311 l ë/kg. The extent of water binding becomes progressively greater as mashes<br />
become more concentrated <strong>and</strong> there is insufficient free water to permit the gelatinization<br />
of much of the starch. The addition of more enzymes to a very thick mash does not<br />
quickly convert the ungelatinized starch <strong>and</strong> so does not enhance the extract obtained.<br />
The situation with the maize starch (Fig. 4.14) is complicated because its gelatinization<br />
temperature (70 75 ëC; 158 167 ëF) is above that of the mashing temperature (65 ëC;<br />
149 ëF) <strong>and</strong> so the conversion of the starch into extract is relatively slow. The potato<br />
starch had a wide gelatinization temperature range (56 69 ëC; 132.8 156.2 ëF), which<br />
spanned the temperature of the mash, <strong>and</strong> the pattern of extract recovery was different<br />
again (Fig. 4.14).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
4.3.8 Wort separation <strong>and</strong> sparging<br />
At the end of the mash the wort is separated from the residual solids. This may be arapid<br />
process, asinmashfilters,oritmaytake1.5 2.5hinsomelautertunsor4 18hinmash<br />
tuns. An extended run-off period allows residual enzymes to continue acting for at least<br />
part of the time. When amash tun or lauter tun is used, the first wort to emerge is diluted<br />
with the water that was originally under the plates. The first runnings are generally<br />
returned to the top of the mash <strong>and</strong> the wort is recycled until it is completely clear <strong>and</strong><br />
`runs bright'. Then wort collection begins, the strong wort emerges <strong>and</strong> gradually the<br />
mash settles onto the plates. Sparging is started <strong>and</strong> the liquor, sprayed onto the surface<br />
fromrotatingsparge-arms,permeatesdownthroughthegoods,progressivelyleachingout<br />
<strong>and</strong> carrying away the remaining extract. In mash tuns this raises the temperature, so the<br />
temperature of the final wort is about 74ëC (165ëF). In contrast, in two- <strong>and</strong> three-vessel<br />
mashing systems the temperature of the whole mash is usually raised to the sparging<br />
temperature <strong>and</strong> after wort recirculation (if this isused), the first wort is collected <strong>and</strong> the<br />
sparge liquor is applied at the same temperature (e.g. 75 78ëC; 167 172.4ëF).<br />
Although sparging temperatures of up to 80ëC (176ëF) may be used, <strong>and</strong> the use of<br />
even higher temperatures has been proposed, these are usually avoided because<br />
undesirable flavours <strong>and</strong> unwanted substances, such as undegraded starch <strong>and</strong><br />
hemicelluloses, may be eluted from the goods. This is particularly likely if undermodified<br />
malt or raw cereal adjuncts have been used. At these elevated temperatures<br />
enzyme destruction is rapid, the rates of diffusion of extract materials from the grist<br />
particles is rapid, the rate of wort separation (`filtration') occurs faster, more protein<br />
aggregation occurs <strong>and</strong> wort viscosity is reduced.<br />
As run off progresses the quality <strong>and</strong> concentration of the wort declines. The last<br />
runnings contain extract that has acomparatively poor quality (Hind, 1950; Figs 4.15,<br />
4.16).Relativetotheextractmorehigh-<strong>and</strong>low-molecularweightnitrogenousmaterials,<br />
ash (including phosphates), silicates (mostly from the silica in the malt husk),<br />
polyphenols <strong>and</strong> astringent substances are dissolved, all these being favoured by the<br />
increasing pH. The specific gravity of the wort rises then declines as the sparge liquor<br />
emerges. As the wort is diluted the fermentability initially increases <strong>and</strong> finally falls<br />
sharply. Often the pH rises, (e.g. by 0.2 0.7), as the buffering substances are eluted from<br />
the goods. The rise is particularly marked if abicarbonate sparge liquor is used. This rise<br />
is undesirable <strong>and</strong> should be checked <strong>and</strong> the calcium ion concentration of the liquor<br />
should be maintained (Laing <strong>and</strong> Taylor, 1984). Experimental thick mashes (liquor/grist<br />
2.5/1, i.e. 28.6%) would not run off unless ahigh concentration of calcium ions (200mg/l)<br />
was used. Thus the last worts are weak, <strong>and</strong> are relatively rich in poorly flavoured<br />
extractives <strong>and</strong> potential haze-forming substances. These last runnings, like the press<br />
liquor from the spent grains (Chapter 3), may be stored hot for ashort period (to prevent<br />
spoilage by micro-organisms) <strong>and</strong> then be added to a subsequent mash to recover the<br />
extract. However, to maintain the quality of the beer the weak wort may need to be<br />
clarified by centrifugation to remove suspended solids (particularly lipids) <strong>and</strong>/or may be<br />
treated with active charcoal (doses of 10 50 g/hl have been suggested) to reduce the<br />
levels of tannins, nitrogenous substances, colour <strong>and</strong> harsh flavours before it is added to a<br />
later mash (Morraye, 1938; Prechtl, 1967).<br />
The faster the wort is run off the higher its content of suspended solids, lipids (which<br />
favour flavour instability, i.e. beer staling), <strong>and</strong> -<strong>and</strong> -glucans which may give<br />
problems when the beer is filtered (Muts <strong>and</strong> Pesman, 1986; Whitear et al., 1983). The<br />
lipid contents of strong worts, separated in various devices, were in increasing order mash<br />
tun < lauter tun < Strainmaster < older pattern mash-filter, <strong>and</strong> were given as 10 < 50<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Contents (mg/100 g of extract)<br />
1400<br />
1000<br />
500<br />
200<br />
a<br />
b<br />
c<br />
d<br />
e<br />
Fermentation<br />
limit<br />
Maltose<br />
19 15 10 5 4 3 2 1 0<br />
Extract in wort (% of dry matter)<br />
impossible without extra manipulations, such as under-letting) in any equipment except<br />
mash filters.<br />
Rates of wort separation are faster with more coarsely ground grists but with these<br />
extract recoveries are less good than from finely ground grists (Section 6.9). The more a<br />
mash is stirred the more fines are produced, the more oxidations are likely to occur<br />
(including the cross-linking of gel-proteins), but if stirring is inadequate temperature<br />
gradients may occur <strong>and</strong> mash may settle <strong>and</strong> burn onto the containing vessel's heating<br />
surfaces, so there is a critical, `compromise' stirring speed (Laing <strong>and</strong> Taylor, 1984). The<br />
deeper the mash-bed the slower wort or sparge liquor will flow through it. Progressively<br />
shallower beds <strong>and</strong> more finely ground grists are used in mash tuns, in lauter tuns <strong>and</strong> in<br />
mash filters. Run-off is impeded by fine particles in the grist (from the malt, cereal flours<br />
or formed in the mash) <strong>and</strong> it is favoured by keeping the malt husk as intact as possible to<br />
give the mash bed a more `open' structure. Mashing under nitrogen gas, experimentally<br />
adding bisulphite (which is a reducing agent), adding heat-stable cellulase, maintaining<br />
adequate levels of calcium ions (particularly in thick mashes), using well-modified malt,<br />
experimentally adding cationic poly-electrolyte flocculants (such as boiled or unboiled<br />
lysozymze or partly de-acetylated chitin) <strong>and</strong> collecting wort at elevated temperatures all<br />
favour rapid wort run-off. Malt may contain endogenous flocculants <strong>and</strong> others may be<br />
present from the fungi present on the surface of the grains (Anderson, 1993). In contrast,<br />
mashing under air or oxygen gas, experimentally adding the oxidizing agent potassium<br />
bromate, omitting calcium ions <strong>and</strong> using poorly modified malt, all favour slow wort runoff<br />
(Anderson, 1993; Barrett et al., 1973, 1975; Crabb <strong>and</strong> Bathgate, 1973; Laing <strong>and</strong><br />
Taylor, 1984; Muller, 1995; Muts et al., 1984).<br />
Poorly modified malts are rich in non-starch polysaccharides (NSPs; pentosans <strong>and</strong> -<br />
glucans) <strong>and</strong> undegraded proteins which are rich in thiol groups. The oxidation of<br />
cysteine side chains produces disulphide links between protein chains that can produce an<br />
insoluble, jelly-like mass of `gel-protein'. Reduction of this material should disperse the<br />
protein gel. An inverse correlation has been established between the gel-protein content<br />
of malt <strong>and</strong> wort separation rate. The hemicellulosic polysaccharides may also form a gel,<br />
which can be attacked by -glucanases, <strong>and</strong> so improve wort separation (Crabb <strong>and</strong><br />
Bathgate, 1973). Fine aggregates of protein, small starch granules, cell-wall non-starchy<br />
polysaccharides (NSPs) <strong>and</strong> lipids can form `high flow-resistant' layers in a mash <strong>and</strong><br />
particularly in the poorly permeable gel-like layer (the Oberteig) which forms on the<br />
surface of mashes in lauter tuns. Removal of this layer reduces the pressure differential<br />
across the bed, increases the flow-rate of the wort but reduces extract recovery (Muts <strong>and</strong><br />
Pesman, 1986). The composition of this layer is variable; two examples contained,<br />
respectively, 18 <strong>and</strong> 20% protein <strong>and</strong> 65 <strong>and</strong> 79% polysaccharides. The composition of<br />
the small aggregates which form in mashing is also variable; for example, small starch<br />
granules, 4 21%, -glucan, 3 19%, pentosan, 5 31% <strong>and</strong> protein 26 42%. In the case<br />
of particles from an all-malt mash, which contained 29% starch, the free lipid content was<br />
5% <strong>and</strong> the bound lipid was 17% (Barrett et al., 1975). The bound lipid may have been<br />
associated with the starch.<br />
These particles probably contribute to, or constitute, the Oberteig. The formation of<br />
this material is favoured by oxidizing conditions. Small starch granules are often firmly<br />
invested with protein which, when oxidized, presumably firmly binds them into the<br />
particles. During mashing the greater part of the malt is dissolved, <strong>and</strong> some proteins are<br />
dissolved <strong>and</strong>, particularly as the temperature rises in temperature-programmed or<br />
decoction mashes, a proportion of the protein is denatured, aggregates <strong>and</strong> precipitates<br />
(Lewis <strong>and</strong> Oh, 1985; BuÈhler et al.,1996). The finer particles (< 1 150 m) tend to block<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the pores of the mash <strong>and</strong> impede run-off, but as the particles aggregate <strong>and</strong> enlarge so<br />
theirobstructiveeffectbecomesless.Largerparticlesarefavouredbycationicflocculants<br />
<strong>and</strong> (apparently) adequate concentrations of calcium ions. Aggregation is better at higher<br />
temperatures, <strong>and</strong> so in three mashes that had been not been heated above 65ëC (149ëF)<br />
offered 3, 3or 3.7 times the specific resistance to the flow of the wort offered by mashes<br />
that had been heated to 80ëC (176ëF; BuÈhler et al., 1996).<br />
`Models' of liquid flow through mashes, for instance based on Poiseuille's equation,<br />
which relates to flow through parallel capillary tubes, or the Carman-Kozeny equation or<br />
Darcy's law, that relate to beds of spherical particles, all emphasize that the rate of flow<br />
through abed of particles is proportional to aconstant, the pressure difference across the<br />
bed, the channel radius to the power 4, to the diameter of the particles squared, <strong>and</strong><br />
inversely proportionaltothedepthofthebed<strong>and</strong>tothe viscosity oftheliquid(Anderson,<br />
1993; Bathgate,1974; Huite <strong>and</strong> Westermann,1974; Laing <strong>and</strong> Taylor, 1984; Meddings<br />
<strong>and</strong> Potter, 1971; Webster, 1978).<br />
As the temperature is increased so wort viscosity falls to comparatively low levels <strong>and</strong><br />
the small particles aggregate <strong>and</strong> increase in size. Both changes favour faster wort<br />
separation. While the viscosity of wort (caused mainly by dissolved sugars <strong>and</strong>, to<br />
varying extents, by polysaccharides <strong>and</strong> perhaps other materials) is not unimportant the<br />
major limitation in wort separation for abed of agiven depth is the `average' particle<br />
diameter, d. Because the flow rate is proportional to d 2 ,as dbecomes smaller so the flow<br />
rate rapidly declines (dˆ1, flowˆ1; dˆ0.5, flowˆ0.25; dˆ0.1, flowˆ0.01, etc.).<br />
Determining an `average' diameter, dis impracticable for the particles in amash, <strong>and</strong> in<br />
any case it is not the depth of the entire mash but the characteristics <strong>and</strong> depth of the<br />
surface Oberteig layer that are often limiting. By using aderived formula that relates to<br />
compressible beds it is possible to find bed permeabilities <strong>and</strong> so test the factors that may<br />
influence them (Laing <strong>and</strong> Taylor, 1984).<br />
4.4 Mashing biochemistry<br />
4.4.1 Wort carbohydrates<br />
The complex mixture of carbohydrates in wort makes up about 92% of the solids in<br />
solution. The most important sugars <strong>and</strong> dextrins in wort are made of glucose, which also<br />
occurs free (4.1). Thus maltose (4.4), maltotriose (4.5) maltotetraose (4.6) <strong>and</strong><br />
maltopentaose (4.7) are made of D-glucopyranose units joined by -(1,4) links (Table<br />
4.15). In cellobiose (4.18) <strong>and</strong> laminaribiose (4.19) the glucose residues are linked by -<br />
(1,4) <strong>and</strong> -(1,3) bonds, respectively. (An introduction to carbohydrate chemistry is<br />
given by Coultate, 2002). The precise composition of the mixture will depend on the<br />
make-up of the grist <strong>and</strong> the mashing conditions. In some `conventional' sweet worts the<br />
carbohydrate spectra are surprisingly similar, whether or not mash tun adjuncts are used<br />
(MacWilliam, 1968). The exceptions are when mashing is carried out to produce `lowalcohol'<br />
beers or `low-carbohydrate' beers when the conditions are chosen to obtain<br />
poorly fermentable <strong>and</strong> maximally fermentable worts, respectively (Chapter 15). Wort<br />
fermentability may be increased by adding amyloglucosidase or, preferably, other<br />
microbial enzymes, such as pullulanase <strong>and</strong> -amylase or fungal saccharogenic amylase<br />
to the mash <strong>and</strong>/or to the fermenter.<br />
The conversion of barley into malt involves aconsiderable loss of potential extract<br />
(Briggs, 1998; Table 4.16). Sugars fermentable by most yeasts are the monosaccharides<br />
glucose (4.1) <strong>and</strong> fructose (4.2), the disaccharides sucrose (4.3) <strong>and</strong> maltose (4.4) <strong>and</strong> the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.15 The major wort carbohydrate fractions compared with the `potentially extractable'<br />
carbohydrates of malt (data of Hall et al., 1956). The brewery extract of the malt was 102.2 lb/Qr<br />
(about 77.5%; 303.5 lë/kg). The laboratory extract was 104.6 lb/Qr (about 78.9%; 309.1 lë/kg). The<br />
values are given as hexose equivalents, that is, as if each fraction had been fully hydrolysed to yield<br />
its component hexoses, <strong>and</strong> so the reported weights are greater than the weights of the unhydrolysed<br />
materials. The carbohydrates made up 91.8% of the wort solids<br />
Malt carbohydrates (hexose equivalents as Wort carbohydrates (hexose equivalents as<br />
% total wort solids) % wort solids)<br />
Starch 85.8 Dextrins, glucans <strong>and</strong> pentosans 22.2<br />
Glucans <strong>and</strong> pentosans* 2.5 ?<br />
Fructans y 1.4 ?<br />
± ± Maltotetraose 6.1<br />
Maltotriose 0.6 Maltotriose 14.0<br />
Maltose 1.0 Maltose 41.1<br />
Sucrose 5.1 Sucrose 5.5<br />
Glucose 1.7 Glucose + Fructose 8.9<br />
Fructose 0.7<br />
Total 98.8 97.8<br />
* These `gums' were soluble in water at 40 ëC (104 ëF). The maltose fraction in the wort contained a trace of<br />
unfermentable isomaltose. y Fructans in the wort were included in the other fractions but were certainly present.<br />
Maltotetraose was essentially absent from the malt.<br />
trisaccharide maltotriose (4.5). Typically maltose is the most abundant sugar in wort.<br />
Sugars in a 12% wort, in g/100 ml, were glucose + fructose, 0.9 1.2; sucrose, 0.4 0.5;<br />
maltose, 5.6 5.9; <strong>and</strong> maltotriose, 1.4 1.7; total 8.3 9.3 (Evans et al., 2002). Some<br />
yeasts only attack maltotriose (4.5) to a limited extent, while other `super-attenuating'<br />
strains may also utilize maltotetraose (4.6) <strong>and</strong> dextrins. The monosaccharides are<br />
Table 4.16 The carbohydrate composition (% dry basis) of Carlsberg barley (TN 1.43%) <strong>and</strong> a<br />
floor-malt made from it (recalculated from Hall et al., 1956). The supposed `structural<br />
carbohydrates' that are not involved with extract formation have been ignored<br />
Barley Malt<br />
Glucose 0.04 1.31<br />
Fructose 0.07 0.55<br />
Sucrose 0.77 3.73<br />
Maltose 0 0.73<br />
Maltotriose 0 0.42<br />
`Glucodifructose' 0.08 0<br />
Raffinose 0.15 0<br />
Fructans 0.58 1.00<br />
Glucans <strong>and</strong> pentosans* 2.10 2.45<br />
Starch 65.86 58.90<br />
Total 69.65 69.09<br />
* Non-starch polysaccharides soluble in warm water at 40 ëC (104 ëF). The raffinose went during<br />
germination, but the `glucodifructose' could not be determined in the malt, <strong>and</strong> will have been<br />
included in another fraction.<br />
The thous<strong>and</strong> corn dry weights of the barley <strong>and</strong> the malt were 39.1 g <strong>and</strong> 35.3 g respectively, so the malt yield<br />
was 35.3 100/39.1 ˆ 90.3%.<br />
100 g barley contained 65.86 g starch hexose, while 90.3 g malt (from 100 g barley) contained 58.90 0.903 g<br />
starch hexose ˆ 53.19 g. Thus the recovery of barley starch in the malt was 53.19 65.68 100 ˆ 80.8%. Thus<br />
the starch going during germination was 19.2%.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO·CH 2<br />
H<br />
H<br />
O<br />
OH<br />
CH2OH H O OH<br />
4<br />
HO<br />
OH H<br />
Fructose; β-Dfructofuranose<br />
(4.2)<br />
6<br />
5<br />
H<br />
OH<br />
H<br />
H OH<br />
3<br />
H<br />
2<br />
*<br />
1<br />
CHO<br />
* 1<br />
H C OH<br />
HO C H<br />
H C OH<br />
H C OH<br />
CH 2OH<br />
β-D-Glucopyranose D-Glucose,<br />
aldose form<br />
OH<br />
CH 2OH<br />
H<br />
CH2OH O H HO·CH2 O<br />
HO<br />
H<br />
OH<br />
H<br />
H OH<br />
O<br />
2<br />
3<br />
4<br />
5<br />
6<br />
(4.1)<br />
Sucrose; α-D-glucopyranosyl-(1→2)β-D-fructofuranoside<br />
CH2OH H O H<br />
HO<br />
H<br />
OH<br />
H<br />
H OH<br />
(4.3)<br />
O<br />
H<br />
CH2OH H O H<br />
HO<br />
OH<br />
H<br />
OH<br />
H OH<br />
OH<br />
H<br />
H OH<br />
H<br />
CH 2OH<br />
H<br />
CH2OH O<br />
*<br />
H, OH<br />
Maltose; α-D-glucopyranosyl-(1→4)-<br />
D-glucopyranose<br />
(4.4)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H<br />
OH<br />
H<br />
H<br />
O<br />
CH2OH H O H<br />
H<br />
OH<br />
H<br />
H OH<br />
CH2OH H O H<br />
4<br />
HO<br />
6<br />
H<br />
5<br />
OH<br />
H OH<br />
3<br />
H<br />
OH<br />
α-D-Glucopyranose<br />
Maltotriose (n = 1); maltotetraose (n = 2); maltopentaose (n = 3)<br />
n<br />
O<br />
2<br />
*<br />
1<br />
CH2OH H O<br />
H<br />
OH<br />
H<br />
H OH<br />
(4.5) (4.6) (4.7)<br />
CH2OH H O H<br />
HO<br />
H<br />
OH OH<br />
H H<br />
HO<br />
CH2OH O H<br />
H<br />
OH H *<br />
H OH<br />
H OH<br />
(4.9)<br />
H, OH *<br />
Mannose; α-Dmanmopyramose<br />
(4.8)<br />
*<br />
OH<br />
Galactose; α-Dgalactopyranose
H<br />
5<br />
H O OH<br />
H<br />
6 OH<br />
HO<br />
H *<br />
H<br />
H OH<br />
3<br />
2<br />
Xylose; β-Dxylopyranose<br />
(4.10)<br />
1<br />
HOCH2 5<br />
H O OH<br />
4<br />
OH H *<br />
H<br />
H OH<br />
3<br />
Arabinose; α-Larabinofuranose<br />
(4.11)<br />
fermented the most rapidly, while maltotriose is fermented slowly <strong>and</strong> sometimes<br />
incompletely so traces may remain in beer. Dextrins, derived from the partial<br />
degradation of starch, are not fermentable <strong>and</strong> neither are pentosans nor are -glucans.<br />
Sometimes the `fermentable sugars' <strong>and</strong> `dextrins' groups are determined <strong>and</strong> the results<br />
are used to calculate the carbohydrate fermentability of the wort as the fermentable<br />
carbohydrates as apercentage of the total carbohydrates. Values in the range 64 77%<br />
are common. The fermentable carbohydrates are the major energy source of the yeast<br />
<strong>and</strong> alcohol <strong>and</strong> carbon dioxide are the major metabolic products. The major source of<br />
extract in an all-malt wort is starch, but preformed sugars are also important <strong>and</strong>of these<br />
sucrose(4.3)isthemostabundant(Tables4.15,4.16).Mannose(4.8)<strong>and</strong>galactose(4.9)<br />
occur combined in malt but neither is released during mashing. In worts maltose (4.4),<br />
along with many other substances, is produced during mashing by the partial hydrolysis<br />
of starch.<br />
Table 4.15, which quantifies the major groups of carbohydrates, indicates the major<br />
sources of the carbohydrates in the extract. However, depending on the grist, isothermal<br />
mashes may yield about 2% of extract from non-starch polysaccharides while this value<br />
may reach 6% in decoction mashes. The value is likely to be higher when the mashes are<br />
supplemented with microbial enzymes, which attack non-starch polysaccharides. The<br />
levels of the monosaccharides in wort should not be abnormally high (as can be the case<br />
when amyloglucosidase is added to the mash) because this can interfere with the uptake<br />
<strong>and</strong> metabolism of maltose (4.4) <strong>and</strong> maltotriose (4.5) by yeast, causing a `sticking<br />
fermentation' (the premature cessation of the fermentation). While the major simple<br />
sugars are as indicated other sugars occur in minor amounts. Pentoses are present (e.g.<br />
xylose (4.10) at 1.5 mg/100 ml, arabinose (4.11) at 1.4 mg/100 ml <strong>and</strong> ribose (4.12) at<br />
0.2 mg/100 ml) compared to maltose (4.4) at 4000 to 6000 mg/100 ml.<br />
The pentoses are more abundant in decoction <strong>and</strong> (probably) temperature-programmed<br />
worts than in simple infusion worts, since the conditions in infusion mashes do not favour<br />
the enzyme-catalysed breakdown of pentosans. Other sugars detected in tiny amounts are<br />
isomaltose (4.13), panose (4.14), isopanose (4.15), nigerose(4.16) <strong>and</strong> maltulose(4.17).<br />
Apparently cellobiose (4.18) <strong>and</strong> laminaribiose (4.19), expected breakdown products of<br />
-glucans, have not been detected. -Glucans <strong>and</strong> pentosans are always present, but in<br />
varying amounts. Carbohydrates also occur in glycolipids, in nucleic acids <strong>and</strong><br />
nucleotides as well as glycoproteins. These are important in various ways, but are<br />
insignificant in terms of extract yield. The calorific value of beer is due to the<br />
unfermentable carbohydrates <strong>and</strong> the ethanol (ethyl alcohol). Small amounts of fructans<br />
occur in malt. These can be regarded as sucrose molecules (4.3) to which one or more<br />
fructose residues have been attached. Simple examples are kestose (4.20), isokestose<br />
(4.21) <strong>and</strong> bifurcose (4.22). The fate of fructans in mashing is not known.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
2<br />
1<br />
HO·CH 2<br />
O OH<br />
H H *<br />
H H<br />
OH<br />
Ribose; β-Dribofuranose<br />
(4.12)<br />
OH<br />
1
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH2OH H OH<br />
H<br />
H<br />
OH<br />
O<br />
H<br />
O<br />
OH<br />
H<br />
H<br />
H, OH *<br />
HO H H O<br />
H OH<br />
CH 2OH<br />
Cellobiose; β-D-glucopyranosyl-(1→4)-<br />
D-glucopyranose<br />
(4.18)<br />
H<br />
CH2OH O H<br />
H<br />
HO·CH2 O H<br />
OH<br />
HO<br />
H<br />
O H HO<br />
CH2· O<br />
H OH<br />
OH H<br />
CH2OH H CH2OH H<br />
H<br />
OH<br />
O<br />
H<br />
O<br />
OH<br />
H<br />
H O<br />
HO<br />
H<br />
H<br />
OH<br />
H<br />
H, OH<br />
OH *<br />
Laminaribiose; β-D-glucopyranosyl-(1→3)-<br />
D-glucopyranose<br />
(4.19)<br />
Kestose (6-kestose); α-D-glucopyranosyl-(1→2)-β-Dfructofuranosyl-(6→2)-β-D-fructofuranoside<br />
(4.20)<br />
CH2OH H O H<br />
H<br />
OH H<br />
HO<br />
HO·CH 2<br />
H<br />
H<br />
HO·CH 2<br />
H OH<br />
HO H<br />
O<br />
H<br />
H<br />
HO·CH 2<br />
OH H<br />
HO<br />
CH2OH 4.4.2 Starch degradation in mashing<br />
Starch makes up the greatest proportion of malt, often about 58% (dry basis), <strong>and</strong> is<br />
present in greater proportions in some mash tun adjuncts. The breakdown products of<br />
starch make up most of the extract in worts. In malt the starch is practically confined to<br />
the starchy endosperm where, in the undermodified regions, it is enclosed in the cell<br />
walls. It is often invested with protein, which may impede its breakdown. Starch is<br />
H<br />
O H<br />
iso-Kestose (1-kestose); α-D-glucopyranosyl-(1→2)-β-Dfructofuranosyl-(1→2)-β-D-fructofuranoside<br />
O<br />
O<br />
(4.21)<br />
O<br />
OH<br />
CH2 HO<br />
HO H<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH2·OH H O H<br />
H<br />
OH H<br />
HO<br />
H OH<br />
O<br />
HO·CH 2<br />
Bifurcose, the basic unit of the two series of fructosans.<br />
Extended in direction A, with β (2→1)-linked fructofuranose<br />
units; kritesin (inulin-type) fructosans. Extended in direction<br />
B with β (2→6)-linked fructofuranose units: hordeacin (phlein type)<br />
H 2·C·O·CO·R 1<br />
HO·C·H O<br />
H<br />
O<br />
OH<br />
CH2OH OH H<br />
CH 2<br />
H<br />
+<br />
H2C·O·P·O·CH2·CH2·N(CH3) 3<br />
O –<br />
O H<br />
O<br />
OH H<br />
HO<br />
CH2·O (4.22)<br />
Lysophosphatidylcholine<br />
R 1·CO- is a fatty acid residue<br />
(4.23)<br />
deposited in organelles called amyloplasts, <strong>and</strong> presumably residues of these also<br />
surround it. It occurs as granules. In barley the granules occur in two populations, the<br />
larger Agranules that may have diameters of 22 48 m, <strong>and</strong> Bgranules with diameters<br />
of 1.7 2.5 m. The large granules make up 10 20% of the granules by number but<br />
85 90% by weight. The large granules have alower gelatinization temperature range<br />
than the small granules.<br />
As the sugar concentration of the surrounding liquid increases, so the gelatinization<br />
temperatures of starches increase (Bathgate <strong>and</strong> Palmer, 1972; Briggs, 1978, 1992, 1998;<br />
Eliasson <strong>and</strong> Tatham, 2001; Letters, 1995a, b; Stone, 1996; Tester, 1997). Starches from<br />
other sources may differ significantly, both in size, shape <strong>and</strong> physical properties<br />
(Chapter 2). The major components of the granules are the polysaccharides amylose <strong>and</strong><br />
amylopectin, together called `starch'. However, the granules are not pure starch, but also<br />
contain some protein, ash <strong>and</strong> lipids. Typically amylose makes up 22 26% of the<br />
polysaccharide, the balance being amylopectin. Amylose is amixture of predominantly<br />
linear -(1,4)-linked chains of D-glucopyranose, about 1600 1900 residues long (Fig.<br />
4.17). The presence of occasional branch-points, formed by -(1,6)-links, is indicated by<br />
the incomplete hydrolysis of amylose by -amylase. In solution amylose can retrograde,<br />
that is crystallize <strong>and</strong> separate from solution. This retrograded material is comparatively<br />
resistant to enzymic attack, <strong>and</strong> so will not readily be converted into extract if formed<br />
during mashing. Amylose adopts a helical shape (six glucose residues/turn) <strong>and</strong> inclusion<br />
compounds can be formed with polar lipids or iodine being contained in the helix. The<br />
complex with iodine has a characteristic blue-black colour. The lipid inclusion<br />
H<br />
B<br />
HO·CH 2<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H<br />
O H<br />
OH H<br />
A<br />
HO<br />
CH2OH
complexes, which occur in barley starch, involve mainly lysophosphatidyl choline, LPC<br />
(4.23), <strong>and</strong> are not readily degraded by enzymes. The polar choline moiety projects from<br />
the end of the helix. The presence of lipids slows, or prevents, retrogradation. Other<br />
cerealstarchescontaininclusioncompoundswithfreefattyacids.Thelipidcomplexesdo<br />
not give colours with iodine<strong>and</strong> so incompletely degraded starch that is complexedis not<br />
detected by the iodine test unless the lipid is first removed, for example, with butanol.<br />
Thus the iodine test, as usually applied to samples from the mash or to spent grains, is<br />
unreliable <strong>and</strong> does not detect all undegraded starch. Each amylose molecule (molecular<br />
weight 26 31 10 4 )will have anon-reducing chain end, in which the terminal glucose<br />
residue is unsubstituted on position C-4, (4.1) <strong>and</strong> areducing chain end where the<br />
terminal glucose has afree C-1 position.<br />
Amylopectin is amixture of highly branched molecules, the -(1,4)-linked chains,<br />
around 26 glucose units long on average, are joined through -(1,6) branch-points,<br />
which may number c. 6% of the bonds in the molecules (Fig. 4.17). Molecular weights<br />
may be very high, e.g., 2 10 6 4 10 8 .Amylopectin is less soluble in water than<br />
amylose. The chains may also adopt helical configurations which, with iodine, give a<br />
red-violet colour. Each amylopectin molecule has only one reducing chain end but<br />
numerous non-reducing chain ends (Fig. 4.17). The polysaccharide molecules are<br />
ordered in the starch granules, which have apartly crystalline structure, as shown by Xray<br />
diffraction. In the granules the crystalline regions alternate with the amorphous<br />
regions, which are more easily attacked by enzymes. The amylose molecules are<br />
supposed to be mixed in among the amylopectin, but the arrangements proposed are<br />
tentative (Fig. 4.18; Imberty et al., 1991). The ordered molecular structure of the<br />
granules is also demonstrated by their birefringence. In polarized light the granules<br />
appear to have adark `maltese cross' on alight background. As the starch is heated in<br />
water <strong>and</strong> gelatinization begins so the swelling granules begin to lose their birefringence<br />
<strong>and</strong> the crosses disappear, a fact that allows an estimation of the gelatinization<br />
temperature range of the starch. Gelatinized starch is readily attacked by enzymes, but<br />
this should take place before retrogradation of any of the amylose occurs. The amyloselipid<br />
complexes of some starches are not disrupted in cooking until temperatures of<br />
90 120ëC (194 248ëF) are reached, which may explain the need to cook some<br />
adjuncts at temperatures above the gelatinization temperatures of their starches. On<br />
cooling inclusion complexes can slowly reform, <strong>and</strong> so the enzymic degradation of the<br />
`liberated' amylose should not be delayed.<br />
`Diastase', the mixture of malt enzymes that catalyses the hydrolytic breakdown of<br />
starch, has been studied for many years, but even now there are uncertainties about the<br />
roles of some of the component enzymes (Fig. 4.19; Briggs, 1992, 1998). Of these<br />
enzymes only the activity of -amylase correlates well with the determination of diastatic<br />
power, DP, as it is usually determined. Older studies attributed the conversion of starch<br />
during mashing to the activities of the - <strong>and</strong> -amylases. While these are the enzymes<br />
chiefly involved it seems that, at least in temperature-programmed mashes, other<br />
enzymes play significant roles.<br />
Malt -amylase is a mixture of different molecules (isoenzymes), having slightly<br />
differing properties <strong>and</strong> these are formed during malting; they are essentially absent from<br />
sound, ungerminated barley. The three `classes' observed each contain multiple forms. -<br />
Amylase-I occurs in comparatively small amounts in malt. It is relatively resistant to acid<br />
conditions <strong>and</strong> chelating agents because it binds calcium ions very strongly. It is inhibited<br />
by small amounts of heavy metal ions, such as copper. -Amylase-II is the `classical'<br />
malt enzyme. It is comparatively resistant to heat, particularly in the presence of excess<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 4.17 Idealized diagrams of (a) amylose <strong>and</strong> (b) amylopectin. The chains of D-glucopyranose<br />
residues (hexagons) are joined by -(1, 4)-links in the amylose <strong>and</strong> the short chains of the<br />
amylopectin which are joined together through -(1, 6)-links, which create branch points, in the<br />
amylopectin. The reducing chain ends are marked with an asterisk, while the non-reducing chain<br />
ends are indicated with solid hexagons. While the straight chained amylose molecule has only one<br />
reducing group <strong>and</strong> one non-reducing chain end, the highly branched amylopectin has one reducing<br />
group but numerous non-reducing chain ends per molecule. The dashed line around the amylopectin<br />
indicates the approximate limit of the -limit dextrin remaining after the molecule has been<br />
attacked by pure -amylase. The shortened branches, of two or three glucose residues, are readily<br />
hydrolysed by limit dextrinase or pullulanase to release maltose or maltotriose. The debranched<br />
dextrin can be degraded further by -amylase.<br />
calcium ions, <strong>and</strong> to heavy metal ions, but it is inhibited by calcium-binding `chelating<br />
agents', such as phytic acid. It isnot completely stable in mashes. It has apH optimum of<br />
about5.3,<strong>and</strong>itisunstableatvaluesbelow4.9.` -Amylase-III'isacomplexbetween -<br />
amylase-II <strong>and</strong> another small protein, BASI, (barley amylase/subtilisin inhibitor), which<br />
limits the activity ofthe enzyme.Thiscomplexis probably disrupted at starch conversion<br />
temperatures. The -amylase mixture from malt attacks the -(1,4)-links within the<br />
starch chains, producing arange of products. Attack is slower at the chain-ends <strong>and</strong><br />
ceases near the -(1,6) branch points. The products of extensive -amylolysis include<br />
glucose (4.1), maltose (4.4), <strong>and</strong> a complex mixture of branched <strong>and</strong> unbranched<br />
oligosaccharides <strong>and</strong> dextrins (Fig. 4.19). Because this kind of attack reduces the starchiodine<br />
colour rapidly, but increases the reducing power of the digest comparatively<br />
slowly this enzyme is often referred to as a`dextrinogenic' amylase. It is also able to<br />
liquefy starch gels. This enzyme is capable of degrading intact starch granules as is -<br />
glucosidase. In mashing, -amylase liberates the dextrins that are the substrate for the<br />
`saccharogenic' -amylase.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Amorphous<br />
Crystalline<br />
Amorphous<br />
Crystalline<br />
Amorphous<br />
Fig. 4.18 Adiagram of the ways in which amylopectin molecules may be packed together in a<br />
starch granule to create amorphous <strong>and</strong> crystalline regions, which differ in their susceptibilities to<br />
enzymolysis (after Imberty et al., 1991). Many of the side chains are wound together in double<br />
helices. The single reducing chain end is at the top of the diagram. Linear amylose molecules, some<br />
of which are in the form of helices <strong>and</strong> inclusion compounds with polar lipids are, in some way,<br />
interspersed with the amylopectin in the granules, possibly in the amorphous regions.<br />
-Amylase occurs in barley in insoluble <strong>and</strong> soluble forms, <strong>and</strong> is of value when raw<br />
barley is used as amash tun adjunct. During malting the proportion of the `free', soluble<br />
enzyme increases as does the ease of extraction. Little or no more enzyme is formed<br />
during germination. The soluble enzyme contains multiple forms having various<br />
molecular weights, including dimers in which enzyme is linked to the, enzymically<br />
inactive, protein Zby disulphide bonds <strong>and</strong> in which high proportions of the monomers<br />
seem to be partially proteolytically degraded forms of at least two genetically distinct<br />
isozymes. The insoluble enzyme can be partly released by proteases, such as papain, by<br />
agents which reduce, <strong>and</strong> so break, the disulphide bonds between the enzyme <strong>and</strong> other<br />
insoluble proteins, <strong>and</strong> by `amphipathic' detergents that disrupt hydrophobic bonds<br />
(Buttimer<strong>and</strong>Briggs,2000).Comparedto -amylase, -amylaseisrelativelysensitiveto<br />
heat <strong>and</strong> heavy metal ions <strong>and</strong> is resistant to mild acidity <strong>and</strong> chelating agents.<br />
-Amylase from different barleys differs in its temperature sensitivity. Barleys with<br />
the more stable enzyme give malts which yield the most fermentable worts (Evans et al.,<br />
2002). This enzyme is readily inactivated by chemical agents that react with thiol groups.<br />
It has abroad pH optimum around 5.0 5.3, (but which alters with the buffer used in the<br />
activity measurements), catalyses the hydrolysis of the penultimate -(1,4)-link of the<br />
non-reducing chain ends of amylose <strong>and</strong> amylopectin, with the release of the reducing<br />
disaccharide maltose (4.4), the most abundant sugar in wort (Fig. 4.19). However, the<br />
enzyme willnothydrolyse -(1,4)bondsnear to -(1,6)branch-pointsinamylopectinor<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Phosphorylase<br />
G–G–G–G–G–… + Pi G–1–P + G–G–G–G–…<br />
Non-reducing<br />
chain end<br />
β-Amylase<br />
Non-reducing<br />
chain end<br />
α-Glucosidase<br />
α-Amylase<br />
Inorganic<br />
phosphate<br />
Glucose-1phosphate<br />
Debranching enzyme (limit dextrinase; R-enzyme)<br />
Shortened<br />
chain<br />
G–G–G–G–(G) N–G… + H 2O G–G* + G–G–(G) N–G…<br />
Maltose Shortened<br />
chain end<br />
G–G* or G–G–G* + H2O G* + G* or G* + GG*<br />
Maltose Maltotriose Glucose Glucose Maltose<br />
G* or G–G*<br />
G<br />
G<br />
Isomaltose<br />
Dextrins, etc.<br />
+ H 2O<br />
G–G–G–G–(G) N –G–G–G–G–(G) M –G* + H 2 O<br />
G–G–G–G–G–G<br />
G* + G* or G* + G–G*<br />
Glucose Glucose Maltose<br />
Branched <strong>and</strong> unbranched chains Mixture of sugars,<br />
oligosaccharides <strong>and</strong> dextrins<br />
Transglucosylase reactions<br />
(These may be catalysed by various enzymes, as ‘side reactions’)<br />
G*, G–G*, G–G–G*<br />
G–G–G–G–G–G–G–G*<br />
G–G–G–G<br />
G–G–G–G–G–G–G–G–G–G* + H 2O G–G–G–G–G–G–G–G–G–G*<br />
G–G–G–G<br />
Amylopectin; branched dextrins<br />
G–G*<br />
Maltose<br />
Glucose + various products<br />
G–G–G–G–G*<br />
Amylose; products without<br />
α(1→6)branch-points<br />
Isomaltose<br />
G–G–G–G–G–G–G–G–(G) N –G–G–G* G–G–G–G–G–G–G–G–G–G*<br />
G–G–G–G–G–G–G<br />
Amylose<br />
Amylopectin<br />
Fig. 4.19 A summary of how the enzymes, that together make up diastase, act on the<br />
polysaccharide components of starch or their breakdown products (after Briggs, 1998). G,<br />
glucopyranose residues ± , (1, 4)-link |; (1, 6)-link; *, reducing group.<br />
-D-<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
G*<br />
G
dextrins. Thus if dispersed starch is attacked by -amylase acting on its own amylose<br />
molecules are broken down, with the liberation of maltose, until one of the occasional<br />
branch-points is encountered, while amylopectin is degraded to maltose <strong>and</strong> a -limit<br />
dextrin in which all the non-reducing chain ends are within two or three glucose residues<br />
ofbranchpoints(Fig.4.17).Inmashingafewofthe -(1,6)-linksmaybebrokenbylimit<br />
dextrinase, with the release of maltose (4.4) or maltotriose (4.5), or -glucosidase, with<br />
the release of glucose (4.1), <strong>and</strong> links within the chains can be broken by -amylase. In<br />
each case the effect is to expose anon-reducing chain end that the -amylase can attack.<br />
The extensive degradation of starch that occurs during mashing depends on the concerted<br />
action of the mixture of enzymes present <strong>and</strong> is often limited by enzyme destruction<br />
rather than the absence of substrates for possible enzyme attack.<br />
Phosphorylase, the enzyme that catalyses the cleavage of the terminal -(1,4) links in<br />
non-reducing chain ends with inorganic phosphate to release glucose-1-phosphate, is<br />
present in barley <strong>and</strong> green malt (Fig. 4.19). Apparently its possible role in mashing has<br />
never been investigated. Like -amylase this enzyme can degrade chains until it comes to<br />
a branch point. As wort contains inorganic phosphate this enzyme may be active <strong>and</strong> as<br />
phosphatases are also present the glucose-1-phosphate generated would be hydrolysed to<br />
glucose <strong>and</strong> phosphate, the overall effect being the same as degradation by -glucosidase.<br />
-Glucosidase is present in barley <strong>and</strong> increases in amount during malting (Briggs,<br />
1998). The enzyme seems to have several forms differing in their substrate specificities <strong>and</strong><br />
a proportion is insoluble but still able to catalyse the hydrolysis of small molecules such as<br />
maltose (4.4) (Fig. 4.19). The enzyme has a pH optimum of about 4.6 <strong>and</strong> a temperature<br />
optimum of 40 45 ëC (104 113 ëF). This enzyme is probably active in the early stages of<br />
temperature-programmed mashing. Like -amylase this enzyme attacks starch granules<br />
(Sun <strong>and</strong> Henson, 1990), <strong>and</strong> acts synergistically with -amylase in this respect. This<br />
enzyme attacks maltose (4.4), isomaltose (4.13), oligosaccharides, dextrins <strong>and</strong> starch at<br />
the ends of the non-reducing chains, hydrolysing -(1, 4) links preferentially <strong>and</strong> -(1, 6)<br />
links more slowly (Fig. 4.19). Like some other carbohydrases this enzyme can catalyse<br />
transglucosylation in strong solutions of sugars, generating small amounts of different<br />
materials. The data on the role of this enzyme in mashing is inadequate, but indirect<br />
evidence suggests that its action can be significant in temperature-programmed mashes.<br />
Debranching enzyme catalyses the hydrolysis of -(1, 6) links in amylopectin <strong>and</strong><br />
dextrins. Previously it was thought that two enzymes were involved <strong>and</strong> these were<br />
referred to as limit dextrinase <strong>and</strong> R-enzyme. When the enzyme acts on -limit dextrins<br />
maltose (4.4) <strong>and</strong> maltotriose (4.5), but not glucose (4.1), are released. While the role of<br />
the malt enzyme in brewer's mashes is uncertain the value of using the similar bacterial<br />
enzyme pullulanase in making highly fermentable worts is clear (Enevoldsen, 1975). The<br />
survival of this enzyme in malt is strongly dependent on the kilning conditions. The<br />
enzyme occurs in insoluble <strong>and</strong> soluble forms <strong>and</strong> much of the soluble enzyme is<br />
inhibited by one or more associated proteins. It was thought that as most of the -(1, 6)<br />
links initially present in the mash survived into the wort the activity of debranching<br />
enzyme must have been insignificant (Enevoldsen, 1975). However, more recent work<br />
indicates that in spite of earlier reports the enzyme is at least as stable as -amylase in<br />
mashes <strong>and</strong> its activity may be significant in some circumstances (Bryce et al., 1995;<br />
Sissons, 1996; SjoÈholm et al.,1995; Stenholm <strong>and</strong> Home, 1999; Stenholm et al., 1996).<br />
Clearly, brewers or distillers who require highly fermentable worts, desire high levels of<br />
active debranching enzyme <strong>and</strong> low levels of the inhibitory proteins in their mashes.<br />
Experiments with soluble starch <strong>and</strong> enzyme preparations in buffered solutions are<br />
unrealistic `models' for mashes in that the conditions are significantly different <strong>and</strong> this<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
alters the stabilities of the enzymes <strong>and</strong> indeed takes no account of the insoluble enzymes<br />
present in malt. -Amylase acting alone on starch for an extended period, at various<br />
temperatures around 65ëC (149ëF), produces poorly fermentable worts (about 20%) <strong>and</strong><br />
this result is comparatively temperature insensitive. If the -amylase is supplemented<br />
with increasing amounts of -amylase (probably contaminated with other enzymes) then<br />
wort fermentability increases. However, the higher the temperature the lower the<br />
fermentability. Thus, as expected, wort fermentability is dependent on both the mashing<br />
temperature <strong>and</strong> the amounts of enzymes present in the mash. At higher temperatures the<br />
heat-labile -amylase is destroyed faster <strong>and</strong> so the `dextrins' produced by the -amylase<br />
areless-well`saccharified'bytheother,moreheatlabileenzymes.Thisisconsistentwith<br />
experience with the results of experimental isothermal mashes (Table 4.17; Fig. 4.20).<br />
Alterations in isothermal mashing temperatures <strong>and</strong> durations <strong>and</strong> malt quality influence<br />
extract yield <strong>and</strong> its quality (Fig. 4.9; Tables 4.6; 4.17). The changes occurring in<br />
temperature-programmedmashingare more complex (Stenholmet al.,1996; Gjertsen<strong>and</strong><br />
Hartlev, 1980; Schur et al., 1973; Table 4.18). Thus, during amash programmed with<br />
restsat48ëC,63ëC,72ëC<strong>and</strong>80ëC(118.4,145.4,161.6<strong>and</strong>176ëF)extractrose,butata<br />
decreasing rate, until the temperature rise from 63 to 72ëC when it became constant (Fig.<br />
4.21). Fermentability stopped rising during the 63ëC rest, when about 0.7 of the -<br />
amylase initially present had been destroyed. Destruction was completed as the<br />
temperature rose to 72ëC. -Amylase destruction was not complete until 80ëC had been<br />
reached.<br />
Whilethepatternsofsugarformationduringmashingreflectenzymeactivitiestheydo<br />
not unambiguously demonstrate which enzymes are active. Thus fructose (4.2) may<br />
originate from the hydrolysis of sucrose (4.3) or higher oligosaccharide fructans, or from<br />
free sugar initially present in the grist. Glucose (4.1) occurs in malt <strong>and</strong> during mashing<br />
may be formed by the activity of -amylase, -glucosidase or -glucosidase. Maltose is<br />
undoubtedly mainly formed by -amylase but there may be contributions from<br />
debranching enzyme (acting as a limit dextrinase) <strong>and</strong> -amylase. In temperatureprogrammed<br />
mashes the rise in glucose at low temperatures suggests that -glucosidase<br />
isactive(Schur et al.,1973). Thepatternsofsugarspresent inisothermalmashes madeat<br />
different temperatures indicate that glucose production is maximal at about 57ëC<br />
(approx. 135ëF), which is consistent with -glucosidase activity (Taylor, 1974; Fig.<br />
4.20).<br />
When starch-containing adjuncts are added in increasing amounts to mashes, wort<br />
fermentability usually declines before extract recovery indicating that saccharogenic<br />
activity becomes limiting before dextrinogenic activity, essentially -amylase. However,<br />
increasing the amount of -amylase in these mashes can increase wort fermentability. In<br />
Table 4.17 The influence of isothermally mashing two malts, at three different temperatures, on<br />
wort fermentability <strong>and</strong> content of nitrogenous substances (Hudson, 1975)<br />
Malt 1 (Diastatic power, Malt 2 (Diastatic power,<br />
Mashing temperature 33 ëL; TN, 1.3%) 90 ëL; TN, 1.8%)<br />
ëC ëF Ferm. (%) TSN (mg/100 ml) Ferm. (%) TSN (mg/100 ml)<br />
68 155 72 73 77 96<br />
65.5 150 76 78 86 106<br />
63 145 79 84 88 113<br />
Abbreviations; Ferm., fermentability. TSN, total soluble nitrogen. TN, total nitrogen content of the malt (% on<br />
dry).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Sugar (% w/v)<br />
10<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
TFS<br />
G2<br />
G 3<br />
SG<br />
1050<br />
1046<br />
1042<br />
1038<br />
1034<br />
0<br />
120 125<br />
S<br />
130 135<br />
G4 140 145 150 155<br />
1030<br />
160<br />
Mashing temperature (°F)<br />
50 55 60 65 70<br />
G<br />
Mashing temperature (°C)<br />
Fig. 4.20 The specific gravity of worts (l ÐÐ l SG) <strong>and</strong> the levels of total fermentable sugars<br />
(ut ± ± ± ut TFS) <strong>and</strong> the sugars present in the worts prepared by isothermal 2h. mashes at the<br />
temperatures shown (after Taylor, 1974). Key to the sugars: G, glucose ± ± ± ; G2, maltose<br />
4 ± ± ± 4 ; G 3, maltotriose s ± ± ± s ; S, sucrose r ± ± ± r ; G 4, maltotetraose n ± ± ± n.<br />
Only traces of fructose were detected.<br />
Table 4.18 The influence of three mashing conditions on the -glucan contents of worts prepared<br />
from three malts (Narziss, 1978). In the extended programme the malt grist was mashed in at 35 ëC<br />
(95 ëF). After a 30 min. rest the temperature was raised (during 15 min.) to 50 ëC (122 ëF) <strong>and</strong> this<br />
was held for 30 min. Temperature rests were subsequently at 65 ëC (149 ëF)/30 min., 70 ëC (158 ëF)/<br />
30 min. <strong>and</strong> 75 ëC (167 ëF)/5 min. In between the rests the temperature was increased at 1 ëC<br />
(1.8 ëF)/min. Total time, 180 min. The shortened mashes began at 50 ëC (122 ëF) or 65 ëC (149 ëF),<br />
<strong>and</strong> in each case the rest of the temperature programme was as before, so mashing times were 135<br />
<strong>and</strong> 90 min. respectively<br />
Malt modification Excellent Normal Poor<br />
Fine/coarse extract difference (% EBC) 1.1 2.0 3.8<br />
Endo- -glucanase activity (mPa/s) 0.343 0.315 0.096<br />
Mashing in temperature Duration -Glucan yield (mg/100 g dry wt.)<br />
ëC ëF (min.)<br />
35 95 180 17 58 390<br />
50 122 135 33 82 595<br />
65 149 90 86 231 645<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Specific gravity of wort (SG)
Extract or fermentable sugars (% w/v)<br />
20<br />
15<br />
10<br />
5<br />
48°C<br />
118.4°F<br />
63°C<br />
145.4°F<br />
β-Amylase<br />
72°C<br />
161.6°F<br />
0 0 40 80 120<br />
Mashing time (min.)<br />
part this may be due to the fermentable sugars produced by this enzyme, but the faster<br />
liquefaction of the starch <strong>and</strong> faster production of dextrins produces more accessible<br />
substrate that -amylase <strong>and</strong> other thermolabile enzymes can attack before they are heat<br />
inactivated. The problem with starches with high gelatinization temperatures is that by<br />
the time they have been liquefied at the necessarily high temperatures used, all the<br />
saccharogenic enzymes have been destroyed. Hence the need to cook <strong>and</strong> liquefy such<br />
starches, <strong>and</strong> adjuncts containing them, then to cool the mixture <strong>and</strong> mix it with malt at<br />
temperatures at which saccharification is still possible. The gelatinization temperature of<br />
malt starch can vary by as much as 6 ëC (10.8 ëF; Bourne, 1998). Worts prepared from<br />
these differing malts had varying fermentabilities.<br />
4.4.3 Non-starch polysaccharides in mashing<br />
The non-starch polysaccharides in the grist (NSP) are the fructans, the hemicelluloses, the<br />
gums <strong>and</strong> the holocellulose. Pectins seem to constitute negligible proportions of grist<br />
materials, although small amounts of combined uronic acids are present. Apart from<br />
sucrose (4.3), which seems to undergo little hydrolysis during mashing, the fate of the<br />
fructans is unknown. However, as the levels of fructose do not increase appreciably<br />
during mashing, it is likely that the fructans, which are very soluble, are not hydrolysed<br />
<strong>and</strong> so they remain with the unfermentable carbohydrates. In many plants fructans are<br />
metabolized by transglycosylation reactions. Holocellulose is the polysaccharide material<br />
which remains undissolved after extracting grist with hot water <strong>and</strong> solutions of caustic<br />
alkalis. This fraction comes mainly from the husk in malt, where it is associated with<br />
lignin <strong>and</strong> it makes up about 5% of barley. However, small amounts are found in all parts<br />
of grains. This material is thought not to undergo any alterations in mashing. There is no<br />
80°C<br />
176°F<br />
Extract<br />
Fermentable<br />
sugars<br />
α-Amylase<br />
Fig. 4.21 The changing yields of extract <strong>and</strong> fermentable sugars (glucose, maltose <strong>and</strong><br />
maltotriose) <strong>and</strong> the declining levels of -amylase <strong>and</strong> -amylase during temperature-programmed<br />
mashing (data of Stenholm et al., 1996).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
400<br />
200<br />
0<br />
α-Amylase (U/g)<br />
1000<br />
500<br />
0<br />
β-Amylase (U/g)
evidence that pure cellulose (poly- -(1,4)D-glucan) is present in malt, although some<br />
could be present in the holocellulose fraction, which contains combined glucose (4.1),<br />
mannose (4.8) <strong>and</strong> lesser amounts of galactose (4.9).<br />
The remaining fractions, usually grouped as the gums <strong>and</strong> hemicelluloses, have been<br />
extensively studied (Briggs, 1998; MacGregor, 1990; Fincher, 1992; Letters, 1995a,b;<br />
Han<strong>and</strong>Schwarz,1996).Togethertheymakeupabout10% ofbarley,butduringmalting<br />
the -glucancomponentissubstantiallydegradedwhilethepentosansincrease.Gums are<br />
soluble in water, while the residual hemicelluloses are soluble in hot solutions of caustic<br />
alkali. If the extraction of gums is carried out with water of increasing temperatures the<br />
quantity of gum recovered increases at the expense of the residual hemicellulose. Thus<br />
these fractions are aseries of materials with arange of solubilities. Chemically there are<br />
two major groups of substances in these fractions, the -glucans, that have been<br />
exhaustively studied, <strong>and</strong> the less-studied pentosans. Minor amounts of other<br />
polysaccharides are probably present. These substances occur in the cell walls of barley<br />
<strong>and</strong> malt. The major, but not the only source of these substances, is the cell walls of the<br />
starchy endosperm in malt <strong>and</strong> barley <strong>and</strong> wheat adjuncts. In barley the carbohydrates of<br />
the endosperm cell walls contain about 70 75% -glucan, 20 25% pentosan <strong>and</strong> 2 4%<br />
holocellulose.<br />
During malting the -glucan in the grain is preferentially degraded. In under-modified<br />
malts, chit malts, inhomogeneous malts <strong>and</strong> barley adjuncts the undegraded gums (<strong>and</strong><br />
possibly hemicelluoses) present give rise to production problems if they are not<br />
adequately broken down during mashing. Problems with mashes made with wheat (or<br />
rye,ortriticale)adjunctsareoftencausedbypentosans.Wortseparationmaybeslow,the<br />
wort will be too viscous, extract recovery is likely to be low, beer filtration will be slow<br />
<strong>and</strong> will use large amounts of filter-aid material, the beers may become hazy <strong>and</strong> even<br />
deposit gels. These materials are also deposited if beer is frozen. However, other<br />
polysaccharides, including -glucan dextrins (derived from starch), yeast glycogen <strong>and</strong><br />
cell-wall glucomannans may also be involved (Forage <strong>and</strong> Letters, 1986; Letters,<br />
1995a,b). Many of the problems attributed to -glucans are probably partly due to<br />
pentosan materials (Han <strong>and</strong> Schwarz, 1996). These polysaccharides may also have<br />
beneficial effects on beer qualities, adding to palate-fullness <strong>and</strong> foam stability. Elevated<br />
levels of -glucans indicate that malt is under-modified <strong>and</strong> so, in addition to other<br />
consequences, may lack adequate `nitrogen (protein)-modification' <strong>and</strong> levels of the<br />
enzymes needed in mashing.<br />
-Glucans are families of molecules consisting of linear chains of -D-glucopyranose<br />
units, of various lengths (molecular weights) linked in various ways. These chains are<br />
unsubstituted <strong>and</strong>, despite speculations to the contrary, there is no evidence for crosslinking<br />
via peptides or other materials. The major class in barley contains amixture of<br />
1,3- <strong>and</strong> 1,4-bonds in which about 90% consists of cellotriosyl <strong>and</strong> cellotetraosyl units<br />
(in which the respectively 3<strong>and</strong> 4glucose units are linked by -(1,4)-bonds) are joined<br />
by single -(1,3)-links (Fig. 4.22). In general the frequency of the (1,3)- to (1,4)-links is<br />
about 3 to 7, but this varies. This material resembles the seaweed polysaccharide lichenin.<br />
Some longer runs of -(1, 3)-links do occur <strong>and</strong> sequences of up to 14 -(1, 4)-linkages<br />
have been reported. In addition small amounts of an exclusively -(1, 3)-linked glucan,<br />
which resembles laminarin, are also present in barley. Apparently, its presence in malt<br />
has not been investigated. The chain lengths of the -glucan molecules vary greatly, the<br />
longer chain materials giving rise to more brewing problems <strong>and</strong> very viscous solutions.<br />
The enzymes involved in the hydrolytic breakdown of -glucans during malting are<br />
indicated in Fig. 4.22. In mashing the amount of -glucan that dissolves is increased by<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
H 2O<br />
β-Glucosidases<br />
H2O β-glucan<br />
solubilase<br />
activities<br />
mashing<br />
G–G–(G)···G–G···G···(G)–G–G–G–G···(G)–G–G–<br />
Limited attack<br />
laminarinase<br />
(β(1→3)<br />
glucanase)<br />
Insoluble, mixed link β-glucan<br />
H 2O<br />
Limited attack<br />
microbial<br />
cellulase<br />
(β(1→4)<br />
glucanase)<br />
Investing<br />
protein; crosslinking<br />
peptides?<br />
H 2O<br />
G–G–(G) a···G–G···G···(G) m–G–G–G–G–G···(G) a···G–G···G<br />
Soluble, mixed-link β-glucan<br />
H 2O<br />
Endo-β-glucanase(s)<br />
G–G···(G) p···G–G–G···G<br />
G···G···(G) q···G–G–G<br />
H<br />
G–G–(G)<br />
2O<br />
r···G<br />
Exo-β-glucanase(s) β-Glucan oligosaccharides<br />
(?)<br />
H2O Endo-β-glucanase(s)<br />
G–G or G···G<br />
Cellobiose Laminaribiose<br />
H 2O<br />
β-Glucosidases<br />
G Glucose<br />
Laminarinase<br />
H2O H2O β-Glucosidases<br />
Acidic<br />
carboxypeptidase<br />
Degradation<br />
products<br />
G···G···G···(G) n···G···G···G···G···G···G<br />
Laminarin-like β-glucan<br />
H2O Laminarinase<br />
(β(1→3)glucanase)<br />
G···G···(G) n···G···G<br />
‘Laminarin’ oligosaccharides<br />
Fig. 4.22 A summary of the activities of the enzymes believed to be involved in the hydrolytic breakdown of<br />
-glucans in germinating grain (Briggs, 1998). G, -D-glucopyranose residue; ± -(1, 4)-link; , -(1, 3)-link.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the activity of aheat-stable enzyme (or mixture of enzymes) now termed -glucan<br />
solubilase (Luchsinger et al., 1958; Scott, 1972; MacGregor <strong>and</strong> Yin, 1990; Bamforth et<br />
al., 1997). The nature of this enzyme has not been established, <strong>and</strong> the enzyme activity<br />
has been attributed to avariety of enzymes acting separately or in concert such as -<br />
(1,3)-glucanase, cellulase ( -(1,4)-glucanase),<strong>and</strong>peptidases.Theseactivitiesmayexert<br />
their effects by partial degradation of the glucan chains to shorter, more soluble<br />
fragmentsortotheremovalof`investing'materialsassociated withthecellwallstructure<br />
(such as proteins <strong>and</strong>/or pentosans). The most significant enzymes are the endo- -<br />
glucanases, which degrade the chains by r<strong>and</strong>om attack on the susceptible bonds,<br />
reducing the viscosity of their solutions. The malt endo- -(1,3)-glucanase, sometimes<br />
called laminarinase, is relatively heat stable but will only attack polysaccharides within<br />
sequences of consecutive (1,3)- -links. Two isozymes have been noted, with the same<br />
pH optimum ofabout 5.6. Their significance inmashing isuncertain. Malts contain small<br />
<strong>and</strong> variable amounts of cellulase(s), enzymes that can only attack -glucan with<br />
consecutive runs of (1,4)-bonds. It is probable that much of this enzymic activity<br />
originates in microbes on the malt.<br />
The more important enzyme, with two isozymes, is properly called endo-(1,3; 1,4)-<br />
-glucan 4-glucanohydrolase, but is usually referred to as (malt) -glucanase. These<br />
enzymes have pH optimaof 4.7<strong>and</strong>hydrolyse the -(1,4)-bond adjacent toasubstituting<br />
-(1,3)-link, giving rise to oligosaccharides which, since these are not known to<br />
accumulate during mashing, are presumably rapidly degraded further by -glucosidases.<br />
Only isozyme EII survives kilning <strong>and</strong> remains in malt (Fincher, 1992). This enzyme is<br />
heat labile <strong>and</strong> amounts remaining in pale malts are very variable. Low levels of malt -<br />
glucanase can give rise to aneed to supplement mashes with preparations of bacterial -<br />
glucanases or fungal glucanases. The -glucosidases, which can hydrolyse laminaribiose<br />
(4.19) <strong>and</strong> cellobiose (4.18) as well as other -glucosides <strong>and</strong> which will attack the nonreducing<br />
chain ends of the -glucan chains, seem to occur as isoezymes having<br />
significantlydifferentspecificities.Theirroleinmashingisuninvestigated.Theexistence<br />
of exo- -glucanases, which were postulated to attack the non-reducing chain ends <strong>and</strong><br />
give rise to cellobiose (4.18) <strong>and</strong> laminaribiose (4.19), has not been confirmed. To be<br />
effective in mashing the malt used must have been carefully kilned to allow the survival<br />
of active -glucanase. In many well-modified, strongly cured ale malts little or none of<br />
this enzyme remains. In all-malt mashes, it is not needed. The enzyme is needed when<br />
inhomogeneous or under-modified malt (including chit malt) is used or -glucan-rich<br />
adjuncts are included in the grist. Then, because the enzyme is heat-labile, the mashing<br />
temperature programme must beadjustedtogive theenzymetimetoact.From aseries of<br />
isothermal mashes, made at different temperatures, it is seen that at temperatures above<br />
45ëC (113ëF) increasing amounts of -glucan are extracted into the wort as higher<br />
temperaturesareused(Fig.4.23).Thiseffectisduetothegreatersolubilityoftheglucans<br />
at higher temperatures combined with the earlier destruction of the -glucanase at higher<br />
temperatures.<br />
In commercial, isothermal mashes it is found that even at temperatures of about 65 ëC<br />
(149 ëF), it is beneficial to have appreciable -glucanase activity in the malt if flake barley<br />
is being used as an adjunct even if, as has been estimated, the enzyme activity survives for<br />
only 2 5 min. Steamed flakes are probably the barley adjunct that most readily releases -<br />
glucan during mashing while micronized barley releases least, perhaps because the<br />
polysaccharide is partly degraded by heat during the preparation of the adjunct. Elevated<br />
wort viscosity, which is an indication of faulty wort production, is often used as a warning<br />
that too much -glucan is in the wort, <strong>and</strong> has even been used to `measure' the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
β-Glucan (mg/l)<br />
600<br />
400<br />
200<br />
55°C<br />
40°C<br />
45°C<br />
50°C<br />
52°C<br />
70°C<br />
(158°F)<br />
50°C<br />
(122°F)<br />
0 40 80<br />
(104, 113°F)<br />
40, 45°C<br />
120<br />
Mashing time (min.)<br />
polysaccharide present. This simplistic view is mistaken (Pierce, 1980; Bathgate <strong>and</strong><br />
Dalgliesh, 1975). Many wort components contribute to its viscosity, including dextrins,<br />
pentosans, <strong>and</strong> sugars. The increase of viscosity with increasing -glucan content is not<br />
linear but is more nearly alogarithmic relationship <strong>and</strong> the viscosity contributions of the<br />
wort components are not simply additive. Temperature-programmed mashing is attractive<br />
inthat,byallowinga`rest'atornearthetemperatureoptimumof -glucanaseitsactivityis<br />
favoured(Table4.18).Howeverextendedlow-temperaturerestswillalsoallowotherheatlabile<br />
enzymes, such as phosphatases, glycosidases <strong>and</strong> proteases, to continue acting,<br />
possibly with undesirable consequences, such as unduly elevated levels of TSN.<br />
The fate of the pentosan polysaccharides in mashing has not been adequately<br />
investigated. Pentosans vary in their sizes <strong>and</strong> detailed structures (Briggs, 1998;<br />
Fincher, 1992). However, they are all based on chains of xylose ( -D-xylopyranose<br />
(4.10)) molecules joined through (1, 4)-links. The xylose chains are variously<br />
substituted with arabinose ( -L-arabinofuranose (4.11)) units. A xylose residue may<br />
be unsubstituted, or be substituted on C-2, C-3 or in both positions. The substitutions<br />
occur irregularly along each chain. Arabinoxylans from malt tissues other than the<br />
starchy endosperm may also be substituted with D-glucuronic acid <strong>and</strong> perhaps<br />
galacturonic acid residues. In addition some of the arabinose substituents may have a<br />
single xylose residue attached. Many studies have treated the pentosans as though they<br />
were pure polysaccharides. However, a proportion of the arabinose units are substituted<br />
with phenolic acids, overwhelmingly ferulic acid (4.131), <strong>and</strong> the molecules are<br />
acetylated. It is assumed that the acetyl residues are attached to the xylose units. The<br />
80°C<br />
(176°F)<br />
60°C<br />
(140°F)<br />
(131°F)<br />
55°C<br />
52°C<br />
(126.6°F)<br />
Fig. 4.23 The extraction, in time, of -glucan, ÐÐÐ , <strong>and</strong> the activity of -glucanase, ..........., in<br />
isothermal mashes made at various temperatures (data of Home et al., 1993).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
0<br />
β-Glucanase activity (U/kg)
H 2O<br />
β-Xylopyranosidase<br />
H 2O<br />
X<br />
X<br />
X<br />
X<br />
Exoxylanase<br />
(?)<br />
F<br />
A<br />
A A<br />
X–X–X–X–X–X–X–X–X–X–X–X–X–X<br />
Ac A Ac Ac Ac A A<br />
Acetic H2O F<br />
acid Ferulic<br />
acid<br />
H2O Deacetylase Deferuloylase<br />
A A<br />
X–X–X–X–X–X–X–X–X–X–X–X–X–X–X<br />
A<br />
A A<br />
X–X–X–X–X–X–X–X<br />
Xylan<br />
H 2O<br />
H 2O<br />
Arabinosidase<br />
A Arabinose<br />
Endoxylanase<br />
consequences of these substitutions on the polysaccharides probably include increased<br />
hydrophobicity, with consequent hydrophobic binding between molecules, including<br />
some proteins, <strong>and</strong> decreased solubility.<br />
Esterase enzymes must be present to remove the acylating (acetic acid <strong>and</strong> ferulic<br />
acid) substituents <strong>and</strong> so expose the polysaccharide to attack by carbohydrases. Mixtures<br />
of microbial esterases <strong>and</strong> carbohydrases act synergistically to break down pentosans.<br />
Pentosans bind large amounts of water, <strong>and</strong> it is this characteristic of the hemicellulose,<br />
present in the fine particles, that is believed to contribute to their slowing wort separation<br />
from mashes. The enzymes believed to be involved in the degradation of pentosans<br />
during malting are indicated in Fig. 4.24. The esterases are unstable in buffer solutions<br />
H 2O<br />
F<br />
Arabinoxylan<br />
Endoxylanase<br />
A<br />
H 2O<br />
X–X–X–X–X–X–X<br />
A A<br />
Arabinoxylan<br />
oligosaccharides<br />
H 2O<br />
Arabinosidase<br />
A Arabinose<br />
X–X–X–X–X–X–X<br />
Shorter chains, Xylan oligosaccharides<br />
X–X<br />
Xylobiose<br />
X<br />
Xylose<br />
H 2O<br />
β-Xylopyranosidase<br />
H 2O<br />
X–X–X<br />
Xylotriose<br />
Endoxylanase<br />
H 2O<br />
A<br />
Deferuloylase<br />
Acetic<br />
acid<br />
X–X–X–X–X<br />
A<br />
F<br />
A<br />
Fate ?<br />
A<br />
F<br />
F<br />
A<br />
X–X–X–X–X–X–X<br />
A<br />
F<br />
A<br />
H 2O<br />
A<br />
F<br />
Arabinosidase<br />
A Arabinose<br />
H 2O<br />
Endoxylanase<br />
F<br />
A<br />
X–X–X–X–X + X–X<br />
A<br />
F<br />
A<br />
A<br />
F<br />
Acetylated <strong>and</strong><br />
feruloylated<br />
arabinoxylan<br />
oligomers<br />
Fig. 4.24 A scheme of the activities of the enzymes believed to be involved in the hydrolytic<br />
breakdown of the grain pentosans during malting (Briggs, 1998). X, -D-xylpyranose residues<br />
(4.10) linked (1, 4) in chains. A, -L-arabinofuranose residues (4.11) Ac <strong>and</strong> F, acetyl <strong>and</strong> feruloyl<br />
substituents, acetic acid <strong>and</strong> ferulic acid (4.131) attached to the polysaccharide through ester links.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
over 30ëC (86ëF) (Humberstone <strong>and</strong> Briggs, 1998). However, feruloyl esterase is active<br />
in mashing with atemperature optimum of around 45ëC (113ëF) (McMurrough et al.,<br />
1984, 1996; Narziss et al., 1990).<br />
The ferulic acid (4.131) liberated is apotential anti-oxidant <strong>and</strong> if decarboxylated<br />
during boiling or by bacteria or pof + yeast strains, gives rise to 4-vinyl guaiacol (4.134,<br />
Fig. 4.33 on page 158), astrongly flavoured substance that is undesirable in most beers.<br />
Which other enzymes are involved in pentosan degradation during mashing is not clear,<br />
but the endo-xylanases, which are relatively heat stable, are probably involved <strong>and</strong> the<br />
greater amounts of arabinose (4.11) <strong>and</strong> xylose (4.10) found in temperature-programmed<br />
mashes, relative to isothermal infusion mashes, indicates that heat-labile glycosidases are<br />
active, at least at lower temperatures. Malt contains an inhibitor of xylanase (Debyser et<br />
al., 1997b). Some 70 90% of malt gums are pentosans, the remainder being chiefly -<br />
glucans. Quoted values are not consistent, but malt may contain 6.4 6.9% pentosan, of<br />
which 0.49 0.69% is water-soluble (Debyser et al., 1997a). Some beers contain<br />
0.3 0.5% (w/v) non-starch polysaccharides, (which may comprise 10% of the beer<br />
carbohydrate), of which c. 70% is pentosan, 23% is -glucan <strong>and</strong> 7% is other materials<br />
(Han <strong>and</strong> Schwarz, 1996). In 15 other beers the arabinoxylans were in the range<br />
514 4211mg/l, while -glucans were in the range 0.3 248mg/l (Schwarz <strong>and</strong> Han,<br />
1995).<br />
4.4.4 Proteins, peptides <strong>and</strong> amino acids<br />
In some analytical systems the nitrogen content of amaterial 6.25, (or some other<br />
factor), is reported as the protein content. This is incorrect <strong>and</strong> misleading, since many<br />
substances besides proteins contain nitrogen. Proteins consist primarily of chains of<br />
amino acids joined by peptide links (Figs 4.25 <strong>and</strong> 4.26). Aprotein may consist of one or<br />
more long polypeptide chains which may or may not be covalently cross-linked through<br />
disulphide bonds between two cysteine (4.31) residues (cystine; 4.32). The chain(s) are<br />
folded together in particular ways <strong>and</strong> the biological activities of proteins, such as<br />
enzymes or lectins, depend on the folding being correct, that is the protein is in its<br />
`native' state. If the folding is disrupted, for example by heat, then the biological activity<br />
is lost, the protein is `denatured' <strong>and</strong>, if it was in solution, it may precipitate. Protein<br />
solution followed by thermal denaturation <strong>and</strong> precipitation occurs in temperatureprogrammed<br />
mashing <strong>and</strong> more aggregation <strong>and</strong> precipitation occurs during the hop-boil,<br />
giving the trub. Native proteins may be soluble or insoluble. Proteins may be substituted<br />
with various molecules, such as sugars in the case of glycoproteins or haem or other<br />
prosthetic groups in the cases of some enzymes, such as peroxidase.<br />
The proteins of cereals are often considered in groups defined by their solubilities<br />
(Briggs, 1998). Globulins are soluble in pure water, while both albumins <strong>and</strong> globulins<br />
are soluble in salt solutions. Many enzymes occur in these soluble fractions but insoluble<br />
enzymes also occur in malt. The hordeins are soluble in hot solutions of aqueous<br />
alcohols, <strong>and</strong> their solubility is enhanced if reducing agents are also present (e.g. 70%<br />
ethanol containing 2-mercaptoethanol, or 60% n-propanol containing sodium borohydride).<br />
The glutelins are soluble only in solutions of strong alkalis. The last two fractions<br />
are largely either reserve materials or, in the case of the glutelins, they may have<br />
structural functions. Some cereal proteins have unexpected properties, for example,<br />
solubility in light petroleum (thionins). When degraded by hydrolysis proteins give rise to<br />
peptides, shorter chains of amino acids (Fig. 4.25), <strong>and</strong> eventually free amino acids (Fig.<br />
4.26 (4.24±4.49)).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
R 6<br />
R 1 –CH–COOH<br />
NH 2<br />
Generalized amino acid<br />
R 1 –CH–CO–NH–CH–COOH<br />
NH 2<br />
Generalized dipeptide<br />
Generalized tripeptide<br />
Generalized protein<br />
The amino acids differ in their properties <strong>and</strong>, when joined in peptide chains, their<br />
side-chains largely define the properties of the peptide or protein of which they form a<br />
part. The amino acid cysteine (4.31) is of interest since the thiol (-SH) on the side chain<br />
can be oxidized to give the `di-amino acid' cystine, containing adisulphide link (4.32).<br />
Thus peptide chains can be cross-linked by covalent disulphide bonds formed by<br />
oxidizing cysteine residues in the chains, a fact that is probably important in<br />
polymerizing gel-proteins during mashing (Van den Berg <strong>and</strong> Van Eerde, 1982; Muller,<br />
1995). Conversely thesebondsmay be split by reducing the disulphide bridges topairs of<br />
thiols. The reduction may be brought about with sodium borohydride, or by thioldisulphide<br />
exchange or by cleavage with bisulphite ions. Presumably, the enhanced rate<br />
of wort run-off caused by the experimental addition of bisulphites to mashes is caused by<br />
the altered structure of the proteins in the fine particles or Oberteig following the<br />
breaking of disulphide cross-links.<br />
Thous<strong>and</strong>s of proteins have been detected <strong>and</strong> from the brewing point of view it is<br />
usually most convenient to consider them in groups defined by particular properties.<br />
However, it is sometimes necessary to consider individual proteins, for example, because<br />
of their enzymic capabilities, or because they bind to particular sugars (i.e. they are<br />
lectins) or they bind to lipids or because they are involved in foam formation <strong>and</strong><br />
stability, or because they are involved in binding to polyphenols giving adducts which<br />
can form hazes in beers. These `haze-forming proteins' can be selectively removed from<br />
beer by adsorbtion onto silica hydrogels, or can be selectively degraded by proteolytic<br />
enzymes such aspapain.These proteins <strong>and</strong> polypeptidesappeartobedistinct from those<br />
which add to the `body' of the beer <strong>and</strong> those which help form <strong>and</strong> stabilize foam.<br />
Surprisingly some malt proteins or modified products (protein Z, 40kDa, <strong>and</strong> LTP 1, a<br />
lipid transfer protein, 10kDa, are examples) partly survive mashing <strong>and</strong> boiling <strong>and</strong><br />
appear, sometimes partly modified, in beer.<br />
R 2<br />
R 3 –CH–CO–NH–CH–CO·NH·CH–COOH<br />
NH 2<br />
R 4<br />
R various ( ) n<br />
NH 2–CH–CO– NH–CH–CO –NH–CH–CO–NH–CH–COOH<br />
Fig. 4.25 Generalized formulae of an -amino acid, adipeptide, atripeptide, <strong>and</strong> asection of a<br />
polypeptide chain, as found in proteins <strong>and</strong> their degradation products. The various side-chains, R,<br />
(Fig. 4.26) differ in their reactivities <strong>and</strong> in some cases may be substituted, for example with<br />
carbohydrates.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
R 7<br />
R 5<br />
R 8
(4.24)<br />
(4.25)<br />
(4.26)<br />
(4.27)<br />
(4.28)<br />
(4.29)<br />
(4.30)<br />
(4.31)<br />
(4.32)<br />
(4.33)<br />
(4.34)<br />
(4.35)<br />
(4.36)<br />
(4.37)<br />
CH 3·CH(NH 2)·COOH<br />
α-Alanine*<br />
NH 2·CH 2·CH 2·COOH<br />
β-Alanine<br />
HOOC·(CH 2) 2·CH(NH 2)·COOH<br />
α-Aminoadipic acid<br />
NH 2·(CH 2) 3·COOH<br />
γ-Aminobutyric acid<br />
NH 2<br />
HN<br />
Arginine*<br />
HOOC·CH 2·CH(NH 2)·COOH<br />
Aspartic acid*<br />
NH 2·CO·CH 2·CH(NH 2)·COOH<br />
Asparagine*<br />
HS·CH 2·CH(NH 2)·COOH<br />
Cysteine*<br />
Cystine*<br />
HOOC·(CH 2) 2·CH(NH 2)·COOH<br />
Glutamic acid*<br />
NH 2·CO(CH 2) 2·CH(NH 2)·COOH<br />
Glutamine*<br />
NH 2·CH 2·COOH<br />
Glycine*<br />
C·NH·(CH 2) 2·CH(NH 2)·COOH<br />
S·CH 2·CH(NH 2)·COOH<br />
S·CH 2·CH(NH 2)·COOH<br />
N<br />
HO<br />
CH 2·CH(NH 2)·COOH<br />
NH<br />
Histidine*<br />
COOH<br />
N<br />
H<br />
Hydroxyproline*<br />
* Occurs in proteins.<br />
(4.38)<br />
(4.39)<br />
(4.40)<br />
(4.41)<br />
(4.42)<br />
(4.43)<br />
(4.44)<br />
(4.45)<br />
(4.46)<br />
(4.47)<br />
(4.48)<br />
(4.49)<br />
CH 3·CH 2<br />
Isoleucine*<br />
CH 3<br />
CH 3<br />
CH 3<br />
Leucine*<br />
NH 2·(CH 2) 4·CH(NH 2)·COOH<br />
Lysine*<br />
CH 3·S·(CH 2) 2·CH(NH 2)·COOH<br />
Methionine*<br />
Phenylalanine*<br />
HO·CH 2·CH(NH 2)·COOH<br />
Serine*<br />
CH 3·CH(OH)·CH(NH 2)·COOH<br />
Threonine*<br />
Tryptophan*<br />
Tyrosine*<br />
Valine*<br />
CH·CH 2·CH(NH 2)·COOH<br />
COOH<br />
N<br />
H<br />
Proline*<br />
CH·CH(NH 2)·COOH<br />
CH 2·CH(NH 2)·COOH<br />
COOH<br />
N<br />
H<br />
Pipecolinic acid<br />
(= piperidine-<br />
2-carboxylic acid)<br />
HO CH 2·CH(NH 2)·COOH<br />
CH 3<br />
CH 3<br />
N<br />
H<br />
CH 2·CH(NH 2)·COOH<br />
CH·CH(NH 2)·COOH<br />
Fig. 4.26 The formulae of most common amino acids <strong>and</strong> the imino acids pipecolinic acid,<br />
proline <strong>and</strong> hydroxyproline.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
It was believed that virtually all the soluble nitrogen-containing substances in wort,<br />
prepared by isothermal mashing at about 65 ëC (149 ëF), were preformed in malt. This is<br />
incorrect. The amounts of soluble nitrogen depend on the malt <strong>and</strong> the way that it is<br />
mashed. In mashes made at 65 ëC (149 ëF) about 50% of the total soluble nitrogen (TSN)<br />
<strong>and</strong> 30 50% of the free amino nitrogen (FAN) is formed by enzyme action during<br />
mashing. The mixture of malt enzymes involved in the hydrolytic breakdown of proteins<br />
is complex (Briggs, 1998; Enari, 1986; Enari et al., 1964; Burger <strong>and</strong> Schroeder, 1976;<br />
Mikola et al., 1972; Sopanen et al., 1980). The major endo-peptidases, or proteases attack<br />
polypeptide chains at various particular locations. Both insoluble (`bound') <strong>and</strong> soluble<br />
(`free') enzyme activities have been detected. The most important proteases are thioldependent,<br />
have pH optima in the range 3 6.5 <strong>and</strong> contribute about 90% of the<br />
proteolytic activity. Metalloproteases (pH optima 5 8.5) provide most of the remaining<br />
activity, but serine proteases <strong>and</strong> aspartate proteases are also present.<br />
About 42 soluble endopeptidases have been detected (Zhang <strong>and</strong> Jones, 1995a, b). In<br />
addition 4 5 amino-peptidases (pH optima 5.5 7.3, which attack peptide chains at their<br />
amino-termini) <strong>and</strong> at least four carboxypeptidases (pH optima 4 6, which attack peptide<br />
chains at the carboxyl termini) together with two alkaline peptidases (pH optima 8 10)<br />
are also present. In mashing, the proteases <strong>and</strong> the carboxypeptidases are the most<br />
important enzymes in generating soluble nitrogenous substances. There is normally an<br />
`excess' of carboxypeptidase activity, <strong>and</strong> so the rate-limiting activity is due to limiting<br />
amounts of proteases. At the end of mashing there is always a substantial amount of<br />
protein remaining in the spent grains <strong>and</strong> so a lack of substrate is not what limits the<br />
generation of soluble nitrogen. During mashing ammonium ions are released as well as<br />
peptides <strong>and</strong> amino acids (Jones <strong>and</strong> Pierce, 1967; Pierce, 1982). Presumably the<br />
ammonium ions arise from the hydrolysis of glutamine (4.34) <strong>and</strong> asparagine (4.30) by<br />
amidases. Apparently transaminases are not active during mashing, but some glutamic<br />
acid (4.33) may be enzymically decarboxylated to give -amino butyric acid (4.27). With<br />
such a complex mixture of enzymes involved it is not surprising that alterations in<br />
mashing conditions can have dramatic effects on the patterns of nitrogenous substances<br />
present in the wort, including the proportions of amino acids. The reported temperature<br />
`optima' of permanently soluble nitrogen (PSN), after 15 min., 1h. <strong>and</strong> 3h. mashing, were<br />
61 ëC (141.8 ëF), 58 ëC (136.4 ëF) <strong>and</strong> 53 ëC (127.4 ëF) respectively while the<br />
corresponding values for formol-nitrogen were 59 ëC (138.2 ëF), 52 ëC (125.6 ëF) <strong>and</strong><br />
50 ëC (122 ëF).<br />
Mash tun adjuncts are often used as sources of extract that will act as `nitrogen<br />
diluents', that is, wort prepared using these adjuncts will contain less soluble nitrogen<br />
than an all-malt wort having the same extract content. However, mash tun adjuncts<br />
contribute some nitrogenous substances to the wort. Wheat adjuncts contribute<br />
polypeptide material that favours foam formation <strong>and</strong> stability, as do preparations of<br />
raw barley. Raw barley (like raw wheat) contains proteins that inhibit proteases from malt<br />
<strong>and</strong> some microbial enzymes. Thus the addition of a barley adjunct to a mash can reduce<br />
the level of wort-soluble nitrogen to a disproportionate extent <strong>and</strong> in barley brewing it is<br />
necessary to ensure that a sufficient amounts of a protease is used to ensure that enough<br />
soluble nitrogen is generated during mashing. In one trial the amounts of -amino<br />
nitrogen in worts (as mg N/kg grist) were, with all malt, 949; with 16% malt replacement<br />
with the named adjunct, wheat, 763; rice, 821; maize, 832 (Jones, 1974).<br />
The proteins <strong>and</strong> polypeptides that survive into the beer contribute to the `body' <strong>and</strong><br />
`mouth-feel' of the beer, its foaming characteristics, <strong>and</strong> its susceptibility to haze<br />
formation. The colour of the beer is influenced by Maillard reactions between the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
sugars <strong>and</strong> amino-compounds (including the amino acids) during the hop-boil, which<br />
give rise to coloured <strong>and</strong> flavoured substances. The proportions of the flavoured<br />
fermentation products made by yeast are dependent on the nitrogenous substances that<br />
are present. The rate of wort separation is reduced by the presence of inadequately<br />
degraded `gel proteins' in the mash contributing to the fine particles <strong>and</strong> the Oberteig<br />
whichimpedetheflowofthewortthroughthegoods.Inwortsfromall-maltmashesthe<br />
levels of amino acids are nearly always adequate for good yeast growth. However, in<br />
worts made using mash tun <strong>and</strong>/or copper adjuncts the FAN levels may fall below the<br />
100 140mg/litre level, which is regarded as the minimum needed for trouble-free<br />
fermentations.<br />
4.4.5 Nucleic acids <strong>and</strong> related substances<br />
Nucleic acids make up 0.2 0.3% of malt. About 70% of the nucleic acid is<br />
deoxyribonucleicacid, DNA(4.50)<strong>and</strong>30%ribonucleicacid,RNA(4.51).Theenzymic<br />
hydrolysis ofthenucleic acidsfirst givesriseto nucleotides(base sugar phosphate),<br />
then nucleosides (base sugar +inorganic phosphate) <strong>and</strong> finally free bases <strong>and</strong> sugars.<br />
In addition, the malt contains avariety of other nucleotides <strong>and</strong> derivatives including<br />
ATP (adenosine triphosphate, (4.53)), NAD + (nicotinamide adenine dinucleotide,<br />
(4.54)), UDPG (uridine diphosphate glucose, (4.55)) <strong>and</strong> so on. The nucleic acids are<br />
chains of alternating phosphate <strong>and</strong> sugar (ribose (4.12) or deoxyribose (4.52)) residues,<br />
eachsugarunitbeingsubstitutedwithapurineorapyrimidinebase(Fig.4.27).Together<br />
these substances contribute 8 9% to the total nitrogen content of malt. During mashing<br />
the degradation of the more complex materials appears to be nearly complete, the<br />
products in the wort being free bases (adenine (4.56), guanine (4.57), cytosine (4.58),<br />
uracil (4.59) <strong>and</strong> thymine (4.60)) or breakdown products such as allantoin (4.61),<br />
hypoxanthine (4.62) <strong>and</strong> xanthine (4.63). The related nucleosides (e.g. adenosine,<br />
guanosine) <strong>and</strong> deoxynucleosides, in which the bases are attached to ribose or<br />
deoxyribose, are also present, as are smaller amounts of nucleotides. (Briggs, 1998;<br />
Ziegler <strong>and</strong> Piendl, 1976).<br />
Several nucleases are present in malt <strong>and</strong> these are sufficiently stable to ensure the<br />
complete hydrolysis of the nucleic acids to nucleotides, molecules in which the<br />
nucleoside structures (base-sugar) are phosphorylated (base-sugar-phosphate) on the<br />
sugar residues in various positions. The phosphatase (nucleotidase) enzymes are also<br />
active as dephosphorylation of the nucleotides to nucleosides proceeds during mashing.<br />
The nucleosidases, which hydrolyse nucleosides to their constituent bases <strong>and</strong> sugars, are<br />
evidently more heat labile since highly kilned malts give relatively more nucleosides to<br />
free bases when mashed, compared to lightly kilned malt <strong>and</strong> mashing at higher<br />
temperatures gives worts richer in nucleosides. Some nucleic acid breakdown products<br />
are known to have flavour-enhancing properties, but the amounts reaching beer are so<br />
small that they can have only a marginal effect. Yeast probably uses the free bases to<br />
support growth in the initial stages of fermentation.<br />
4.4.6 Miscellaneous substances containing nitrogen<br />
A wide range of nitrogen-containing substances, besides proteins, peptides <strong>and</strong> amino<br />
acids, occur in malt <strong>and</strong> wort in widely differing amounts (Briggs,1998; Engan, 1981;<br />
MacWilliam, 1968; Fig. 4.28). Ammonia, as ammonium ions, <strong>and</strong> many amines, often<br />
formed by the decarboxylation of amino acids, are found in worts. Thus glycine (4.35)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
B<br />
H<br />
H<br />
B<br />
H<br />
H<br />
B<br />
H<br />
H<br />
O<br />
H2C–O··· x O<br />
H<br />
H<br />
O H2C–O– P O<br />
x O<br />
H<br />
H<br />
O H2C–O– P O<br />
x O<br />
H<br />
H 2C–O– P O<br />
NH 2<br />
OH<br />
C<br />
N C<br />
N<br />
CH<br />
HC C<br />
N N O<br />
OH<br />
H H<br />
H H<br />
OH<br />
OH<br />
HO·CH 2<br />
H<br />
H<br />
O<br />
OH H<br />
OH<br />
H<br />
H<br />
*<br />
Deoxyribose; β-Ddeoxyribofuranose<br />
(4.52)<br />
Deoxyribonucleic acid, DNA<br />
(x = H, Major bases, B: Adenine, guanine, cytosine, thymine)<br />
Ribonucleic acid, RNA<br />
(x = OH, Major bases, B; Adenine, guanine, cytosine, uracil)<br />
OH<br />
O O O<br />
CH 2·O·P·O·P·O·P·O –<br />
O –<br />
O –<br />
Adenosine triphosphate ATP<br />
(4.53)<br />
O –<br />
CH2OH H O<br />
H<br />
OH<br />
HO<br />
H OH<br />
NH 2<br />
O·R<br />
OH<br />
(4.54)<br />
OH<br />
(4.50)<br />
(4.51)<br />
O CH O O<br />
2·O·P·O·P·OCH2 O<br />
+<br />
N<br />
H<br />
H<br />
H<br />
H<br />
O H<br />
H<br />
H<br />
H<br />
– O –<br />
N C<br />
C<br />
N<br />
CH<br />
N<br />
C<br />
HC<br />
N<br />
H<br />
O·P·O·P·O·CH2 O<br />
O<br />
O<br />
N<br />
H<br />
H<br />
H<br />
H<br />
OH<br />
H·N<br />
OH<br />
(4.55) Uridine diphosphate<br />
glucose, UDPG<br />
OH<br />
Nicotinamide adenine.<br />
dinucleotide, NAD +<br />
R = H<br />
NADP + , R = phosphate<br />
CO·NH 3<br />
Fig. 4.27 The structures of the nucleic acids, the purine <strong>and</strong> pyrimidine bases <strong>and</strong> the breakdown<br />
products allantoin, xanthine <strong>and</strong> hypoxanthine <strong>and</strong> of three chemically related cofactors.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 4.27 Continued.<br />
gives rise to methylamine (4.64), alanine (4.24) gives ethylamine (4.65), valine (4.49)<br />
gives isobutylamine (4.66), phenylalanine (4.42) -phenylethylamine (4.67), tyrosine<br />
(4.48) tyramine (4.68), histidine (4.36) histamine (4.69), tryptophan (4.47) tryptamine<br />
(4.70), <strong>and</strong> proline (4.44) pyrrolidine (4.71). Similarly, the decarboxylation of arginine<br />
(4.28) yields agmatine (4.72), while ornithine (4.73) <strong>and</strong> lysine (4.40) give the diamines<br />
putrescine ((4.74); 1, 4-diaminobutane) <strong>and</strong> cadaverine ((4.75) 1, 5 diaminopentane),<br />
respectively. These last are precursors of spermine (4.76) <strong>and</strong> spermidine ((4.77) Briggs,<br />
1978). The origins of dimethylamine (4.78), trimethylamine (4.79) p-hydroxybenzylamine<br />
(4.80) <strong>and</strong> gramine (4.81), butylamine <strong>and</strong> amylamine are less obvious. The<br />
methylated derivatives of tyramine (4.68), N-methyl-tyramine (4.82), di-N-methyl<br />
tyramine ((4.83), hordenine) <strong>and</strong> the quaternary tri-N-methyl-tyramine ((4.84), c<strong>and</strong>icine)<br />
are also present.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Awide range of heterocyclic, N-containing substances also occurs in worts prepared<br />
using dark or roasted malts or adjuncts. Various other N-containing substances are also<br />
vitamins <strong>and</strong>/or yeast growth factors. Other substances, which should be present in only<br />
tiny amounts or be absent, include nitrosoamines <strong>and</strong> hydrocyanic (prussic) acid (Fig.<br />
4.28). Nitrosamines can arise when malt is kilned <strong>and</strong> oxides of nitrogen are present in<br />
the air. Using modern malting techniques these substances should be nearly absent. The<br />
compound originally attracting interest was N-nitrosodimethylamine (NDMA, (4.85)),<br />
which was formed on the surface of the malt, but other substances, such as Nnitrosoproline<br />
have subsequently been detected. Barley malts contain widely variable<br />
amounts of cyanogenic glycosides, such as epi-heterodendrin (4.86). The amounts<br />
formed during malting are strongly influenced by the barley variety. This compound can<br />
be hydrolysed to glucose <strong>and</strong> isobutyraldehyde cyanohydrin by the enzyme -<br />
glucosidase, then the cyanohydrin breaks down to isobutyraldehyde <strong>and</strong> hydrocyanic<br />
acid. Traces of this acid have been reported in beers, but the levels present are<br />
insignificant. This is not the case in distilleries, when the acid can give rise to urethane<br />
(ethyl carbamate, (4.87)). Some sorghum malts contain very large amounts of dhurrin<br />
(4.88), another cyanogenic glycoside, <strong>and</strong> in these cases there is a clear risk that<br />
significant levels of hydrocyanic acid (HCN) may reach beers made from them (Briggs,<br />
1998). As with barley the amount of cyanogenic glycoside present in the sorghum malt is<br />
strongly influenced by the variety.<br />
4.4.7 Vitamins <strong>and</strong> yeast growth factors<br />
Many of the growth factors needed by yeast contain nitrogen (Table 4.19; Fig. 4.29). In<br />
addition to the substances mentioned in this section brewer's yeast needs some sterols<br />
<strong>and</strong> unsaturated fatty acids for growth, after periods of anaerobic growth (Chapter 12).<br />
The quantities of vitamins reported in worts vary widely. In part these discrepancies<br />
probably represent real differences <strong>and</strong> in part are caused by difficulties with the<br />
bioassays used in the estimations. Many of these substances occur combined in various<br />
ways <strong>and</strong> these may or may not be broken down during mashing <strong>and</strong> may or may not be<br />
available to yeast. Thus folic acid (4.89) <strong>and</strong> related compounds occur in various forms,<br />
nicotinic acid occurs (as nicotinamide (4.90)) in the oxidation/reduction cofactors NAD +<br />
(4.54) <strong>and</strong> NADP + ,<strong>and</strong> myo-inositol (4.91) occurs combined with phosphate in phytic<br />
acid (4.156) <strong>and</strong> in some lipids (e.g. (4.122)). Riboflavin (4.92) occurs in the oxidation/<br />
reduction co-factors flavin mononucleotide (FMN, (4.93)) <strong>and</strong> flavin adenine dinucleotide<br />
(FAD, (4.94)), while thiamine (4.95) occurs as thiamine pyrophosphate.<br />
Malt contains some fat-soluble vitamin precursors (Briggs, 1998), of which the<br />
tocopherols (vitamin E) might be significant, but it is unclear if any of these reach the<br />
hoppedwort.Thewatersolublevitaminsaresignificant(Table4.19).AlthoughvitaminC<br />
(ascorbic acid (4.96) <strong>and</strong> dehydroascorbic acid (4.97)) are present in green malt these<br />
materials are destroyed during kilning. Traces of vitamin B12 have sometimes been found<br />
in malt, but the significance is unclear. It is not known what alterations may occur to<br />
vitamins <strong>and</strong> their precursor substances during mashing. At first it seems strange that so<br />
many uncertainties surround these compounds. The reason is probably that the levels<br />
present in conventional worts rarely or never limit yeast growth <strong>and</strong> fermentation, <strong>and</strong> so<br />
there is no stimulus for investigating their origins <strong>and</strong> fates. However, problems with<br />
fermentations have been encountered with high-adjunct, barley brews <strong>and</strong> these have<br />
been overcome by the addition of yeast extracts to the worts. Various water soluble<br />
vitamins <strong>and</strong> growth factors are known to be present in these complex preparations <strong>and</strong> so<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH 3·NH 2<br />
Methylamine<br />
(4.64)<br />
CH3·CH2·NH2 Ethylamine<br />
(4.65)<br />
(CH3) 2CH·CH2·NH2 Isobutylamine<br />
(4.66)<br />
HO CH 2·CH 2·NH 2<br />
N<br />
CH 2·CH 2·NH 2<br />
Phenylethylamine<br />
(4.67)<br />
Tyramine<br />
(4.68)<br />
CH 2·CH 2·NH 2<br />
NH<br />
Histamine<br />
(4.69)<br />
CH 2·CH 2·NH 3<br />
N<br />
H<br />
Tryptamine<br />
(4.70)<br />
N<br />
H<br />
Pyrrolidine<br />
(4.71)<br />
NH 2·(CH 2) 4NH·C<br />
Agmatine<br />
(4.72)<br />
NH2·(CH2) 3·CH(NH2)·COOH Ornithine<br />
(4.73)<br />
NH2·(CH2) 4·NH2 Putrescine<br />
(4.74)<br />
NH2·(CH2) 5·NH2 Cadaverine<br />
(4.75)<br />
NH2·(CH2) 3·NH·(CH2) 4·NH·(CH2) 3·NH2 Spermine<br />
(4.76)<br />
NH<br />
NH 2<br />
NH2·(CH2) 3·NH·(CH2) 4·NH2 Spermidine<br />
(4.77)<br />
HO<br />
CH 2·CH 2·N(CH 3) 2<br />
Hordenine; di-N-methyl tyramine<br />
HO<br />
(CH3) 2·NH<br />
Dimethylamine<br />
(4.78)<br />
Trimethylamine<br />
HO CH 2·NH 2<br />
p-Hydroxybenzylamine<br />
(4.80)<br />
N<br />
H<br />
Gramine<br />
(4.81)<br />
(4.83)<br />
+<br />
CH2·CH2·N(CH3) 3<br />
C<strong>and</strong>icine; tri-N-methyl tyramine<br />
(4.84)<br />
OH –<br />
NDMA; N-Nitroso-dimethylamine<br />
(4.85)<br />
(4.86)<br />
NH2·COO·C2H5 Urethane (ethyl carbamate)<br />
(4.87)<br />
CH 2·N(CH 3) 2<br />
HO CH 2·CH 2·NH·CH 3<br />
N-Methyltyramine<br />
CH 3<br />
CH 3<br />
CH2OH CH(CH3) 2<br />
H O<br />
O C CN<br />
H<br />
OH H<br />
HO H<br />
H<br />
H OH<br />
(CH 3) 3N<br />
(4.79)<br />
(4.82)<br />
N·NO<br />
epi-Heterodendrin<br />
CH2OH H<br />
H O<br />
O H<br />
OH H<br />
HO OH<br />
C<br />
CN<br />
OH<br />
H OH<br />
Dhurrin<br />
(4.88)<br />
Fig. 4.28 Some amines <strong>and</strong> other nitrogenous substances found in wort.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.19 Vitamins <strong>and</strong> yeast growth factors in wort (Briggs et al., 1981; MacWilliam, 1968).<br />
The amounts of these factors present in worts are likely to vary considerably with the different<br />
compositions of grists <strong>and</strong> with differing wort concentrations<br />
Growth factor or vitamin Reported concentrations<br />
Choline (4.101) 20±25 mg/100ml<br />
myo-Inositol (4.91) bound 1±6mg/100ml<br />
free 1.5±3.5 mg/100ml<br />
Thiamine ((4.95); aneurine; vitamin B1) 28±75(155) g/100ml<br />
Riboflavin ((4.92); vitamin B2) 33±90 g/100 ml<br />
Folic acid (4.89), <strong>and</strong> related substances 10±13 g/100 ml<br />
Nicotinic acid ((4.90); niacin) 0.8±1.8 mg/100 ml<br />
Pyridoxin ((4.100); pyridoxal, pyridoxamine, vit. B6) 59±105 g/100 ml<br />
Biotin (4.98) 0.8±1.2 g/100 ml<br />
Pantothenic acid (4.99) 48±98 g/100 ml<br />
these may be partly or wholly responsible for the improved fermentation performance of<br />
the yeast.<br />
4.4.8 Lipids in mashing<br />
Malt contains about 3.5% of lipids <strong>and</strong> is the major source of lipids in all-malt beers.<br />
Most adjuncts contain less lipid than malt, <strong>and</strong> specifications often specify the maximum<br />
that any batch may contain. They also contribute some lipid to the wort. Older reports<br />
often quoted low values for lipid contents, because only the non-polar materials were<br />
extracted by the methods then in use. Malt lipids are acomplex mixture containing<br />
hydrocarbons, fatty acid esters, sterol esters, esterified steryl glycosides, waxy esters,<br />
monoglycerides, diglycerides <strong>and</strong> triglycerides, free sterols, free fatty acids, long chain<br />
alcohols, phospholipids, glycolipids, carotenoids <strong>and</strong> tocopherols (Figs 4.30 <strong>and</strong> 4,31;<br />
Table 4.20; Anness, 1984; Briggs, 1978, 1998; MacWilliam, 1968; Morrison, 1978,<br />
1988). <strong>Brewing</strong> <strong>science</strong> has concentrated on relatively few of these groups of substances.<br />
Using the analysis of fatty acids, the lipid types present in afree wort (as mg fatty<br />
acid/litre) were reported to be: phospholipids+glycolipids 14.8, monoglycerides, 1.7;<br />
diglycerides, 2.8; triglycerides, 15.3; free fatty acids, 28.4; steryl esters, 1.0 <strong>and</strong><br />
unknowns, 0.3 (Table 4.21). These analyses do not record lipids lacking fatty acids in<br />
their make-up. Triglycerides (4.102) predominate in malt, but free fatty acids (4.107±<br />
4.113) predominate in the wort. The amounts of lipid extracted into wort is increased by<br />
using better modified malts, finer grinding, higher mashing <strong>and</strong> sparging temperatures,<br />
thinner mash beds during wort separation, careless <strong>and</strong> excessive raking in the lauter tun,<br />
by using smaller proportions of adjuncts, <strong>and</strong> by running off faster, by `squeezing' the<br />
mash to recover residual extract, or by adopting other techniques to maximize extract<br />
recovery. Much of the lipid in amash is present as oil droplets spread among the grist<br />
particles. In general, more turbid worts carry more lipids <strong>and</strong> techniques are usually<br />
adopted to minimize turbidity <strong>and</strong> the amounts of lipid remaining in the wort, even<br />
though their presence can increase the fermentation rate (Chapter 12). Thus the wort is<br />
recirculated through the filter bed until it `runs bright', or it can be filtered through<br />
kieselguhr or centrifuged. The last two treatments are probably not much used on the<br />
production scale.<br />
More lipids are also removed at later stages of the brewing process, for instance,<br />
during wort boiling <strong>and</strong> clarification, but it is sound <strong>practice</strong> to obtain the sweet worts as<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
H 2N<br />
H 3C<br />
H·N<br />
N<br />
H 3C<br />
H 3C<br />
OH<br />
N<br />
Riboflavine, vitamin B 2<br />
(4.92)<br />
CH 2OH<br />
(CHOH) 3<br />
CH 2<br />
N N<br />
N<br />
CH 2·NH CO·NH·CH·CH 2·CH 2·COOH<br />
Pteridine p-Aminobenzoic<br />
acid residue<br />
N<br />
O<br />
S<br />
NH 3<br />
N<br />
N·H<br />
N<br />
N<br />
(4.98)<br />
Pteroic acid<br />
Cl<br />
CH 2<br />
N<br />
Cl<br />
O<br />
Folic acid<br />
(4.89)<br />
NH<br />
O<br />
Vitamin B 1 (Thiamin, Aneurine)<br />
S<br />
Ascorbic acid<br />
(4.95) (4.96)<br />
(CH2) 4·COOH<br />
Biotin<br />
CH 3<br />
CH 2·CH 2·OH<br />
CH 3<br />
O<br />
COOH<br />
Glutamic acid residue<br />
Flavin mononucleotide;<br />
FMN or riboflavin phosphate<br />
O<br />
H·N<br />
O<br />
OH OH<br />
N<br />
CH 2·OH<br />
CH·OH<br />
(4.93)<br />
N<br />
Nicotinic acid (R = OH)<br />
Nicotinamide (R = NH 2)<br />
CH 3<br />
O N N CH3 CH 2<br />
H·C·OH<br />
H·C·OH<br />
H·C·OH<br />
2–<br />
CH2OPO3 HO·CH 2·C·CHOH·CO·NH·CH 2·CH 2·COOH<br />
CH 3<br />
Pantothenic acid<br />
(4.99)<br />
β-alanine residue<br />
O<br />
O<br />
O O<br />
(4.97)<br />
CH 2·OH<br />
CH·OH<br />
(4.90)<br />
Dehydroascorbic acid<br />
HO<br />
H 3C<br />
R<br />
N<br />
CO·R<br />
H·N<br />
O<br />
Pyridoxin (R = CH 3OH)<br />
Pyridoxal (R = CHO)<br />
Pyridoxamine (R = CH 2·NH 2)<br />
(4.100)<br />
N<br />
CH 2<br />
H·C·OH<br />
H·C·OH<br />
H·C·OH<br />
CH 2<br />
O<br />
– O P O<br />
O<br />
(4.94)<br />
OH OH<br />
H OH<br />
H<br />
OH<br />
H<br />
H<br />
HO H<br />
H OH<br />
myo-inositol<br />
(4.91)<br />
CH 3<br />
O N N CH3 NH 2<br />
–<br />
O O P<br />
O<br />
N N<br />
CH2 O<br />
N N<br />
H<br />
H<br />
H<br />
OH OH<br />
Flavin adenine dinucleotide, FAD<br />
CH 2OH<br />
(CH 3) 3·N·CH 2·CH 2·OH<br />
OH<br />
–<br />
Choline<br />
(4.101)<br />
Fig. 4.29 Some water-soluble vitamins <strong>and</strong> yeast growth factors that occur in wort <strong>and</strong> ascorbic <strong>and</strong><br />
dehydroascorbic acids, which occur in green malt.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
H 2·C·O·CO·R 1<br />
R 2 ·CO·O·C·H<br />
H 2·C·O·CO·R 3<br />
Triglyceride (triacylglycerol)<br />
(R·CO·OH = fatty acid)<br />
HO<br />
H 2·C·O·CO·R<br />
HO·OH<br />
H 2·C·OH<br />
1-Monoglyceride<br />
(monoacylglycerol)<br />
CH 3<br />
CH3 CH3 β-sitosterol<br />
CH 3<br />
H 2·C·O·CO·R 1<br />
R 2 ·CO·O·C·H<br />
H 2·C·OH<br />
1,2-Diglyceride<br />
(diacylglycerol)<br />
H 2·CO·CO·R 1<br />
HO·CH<br />
H 2CO·CO·R 2<br />
1,3-Diglyceride<br />
(diacylglycerol)<br />
(4.102) (4.103) (4.104) (4.105)<br />
bright as possible. Less than 5% of malt lipids are extracted into the sweet wort, the<br />
amount depending on the equipment used <strong>and</strong> the way it is operated. The amounts of malt<br />
lipids in sweet worts (expressed as % of the lipid in the malt) prepared using a traditional<br />
mash tun, a lauter tun <strong>and</strong> a traditional mash filter were 0.3%, 1.0% <strong>and</strong> 4.5%<br />
respectively (Anness <strong>and</strong> Reid, 1985). In contrast, the new 2001 filter releases relatively<br />
little lipid into the wort (Letters, 1994). In another report the lipid contents of worts<br />
prepared using a deep bed mash tun, a lauter tun, a Strainmaster <strong>and</strong> a traditional mash<br />
filter were 10, 50, 150 <strong>and</strong> 400 mg/litre, respectively (Whitear et al., 1983).<br />
Lipids have a number of effects in brewing but there are disagreements about the<br />
details, probably because different criteria <strong>and</strong> lipid fractions have been used in the<br />
studies (Letters, 1992, 1994; Letters et al., 1986; Isherwood et al., 1977; Wainwright,<br />
1980). Unsaturated fatty acids <strong>and</strong> sterols (Fig. 4.30) have a beneficial effect on yeast,<br />
improving its fermentation performance, its viability <strong>and</strong> its resistance to high alcohol<br />
contents, such as occur in high-gravity brewing. These materials can only be produced by<br />
yeast in sufficient amounts under aerobic conditions. When turbid worts, which contain<br />
elevated amounts of lipids, reach the fermenter it is often found that fermentation is<br />
enhanced. It seems that, in the amounts found normally in beers lipids do not influence<br />
gushing. On the other h<strong>and</strong> the lysophospholipids (such as lysophosphatidyl choline,<br />
(4.23)) complexed with amylose in starch <strong>and</strong> the lysophospholipids together with the<br />
free fatty acids present in mashes can slow down starch gelatinization <strong>and</strong> impede<br />
amylolysis <strong>and</strong> the degradation of complexed dextrins. As the lipid-polysaccharide<br />
complexes do not give positive iodine tests these can be misleading when estimating the<br />
amounts of starch remaining in the spent grains.<br />
CH 3<br />
CH 3<br />
(4.106)<br />
Common fatty acids<br />
CH 3·(CH 2) 12·COOH Myristic acid (14:0)<br />
CH 3·(CH 2) 14·COOH Palmitic acid (16:0)<br />
CH 3·(CH 2) 5·CH =CH·(CH 2) 7·COOH Palmitoleic acid (16:1)<br />
CH 3·(CH 2) 16·COOH Stearic acid (18:0)<br />
CH 3·(CH 2) 7·CH =CH·(CH 2) 7·COOH Oleic acid (18:1)<br />
CH 3·(CH 2) 4·CH =CH·CH 2·CH==CH·(CH 2) 7·COOH Linoleic acid (18:2)<br />
CH 3·CH 2·CH =CH·CH 2·CH =CH·CH 2·CH =CH·(CH 2) 7·COOH Linolenic acid (18:3)<br />
(4.107)<br />
(4.108)<br />
(4.109)<br />
(4.110)<br />
(4.111)<br />
(4.112)<br />
(4.113)<br />
Fig. 4.30 Non-polar lipids <strong>and</strong> fatty acids, with -sitosterol as an example of a sterol. R.CO.O<br />
represents an esterified fatty acid residue.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(4.114)<br />
(4.118)<br />
(4.120)<br />
CH 2OH<br />
HO O<br />
H<br />
OH H<br />
H H<br />
(4.116)<br />
H OH<br />
O<br />
CH 2<br />
Mono-β-D-galactosyl<br />
diglyceride<br />
R 2·CO·O·CH<br />
CH·O·CO·R 1<br />
CH 2·O·CO·R 2<br />
H 2C·O·CO·R 1<br />
H 2C·O· P<br />
CH 2OH<br />
OH O H<br />
H<br />
OH H<br />
H<br />
H OH<br />
Phosphatidyl inositol<br />
Digalactosyl diglyceride<br />
The addition of most preparations of lipids to beer reduces head formation <strong>and</strong><br />
survival. The effects are complex, both because different mixtures of lipids have different<br />
effects <strong>and</strong> because lipid binding proteins (such as LTP1, a 10 kDa albumin) can survive<br />
into beer <strong>and</strong> `mask' the effects of the lipids. Indeed, LTP1 constitutes a major proportion<br />
of the protein in foam. At least in some combinations, mixtures of different groups of<br />
lipids act synergistically to destroy foam when added experimentally to beer. Possibly<br />
O<br />
HO<br />
CH 2<br />
O<br />
OH H<br />
H H<br />
H OH<br />
Phosphatidic acid Phosphatidylglycerol<br />
H 2C·O·CO·R 1<br />
R2·CO·O·CH O<br />
+<br />
H2C·O·P·O·CH2·CH2·N(CH3) 3<br />
O –<br />
O<br />
H 2C·O·CO·R 1<br />
R 2·CO·O·CH O<br />
H 2C·O·CO·R 1<br />
R2·CO·O·CH O<br />
H 2C·O·P·O·CH 2<br />
H 2C·O·P·O·CH 2<br />
O – H·C·OH<br />
O – H·C·OH O<br />
CH 2<br />
CH 2OH<br />
R 4·CO·O·CH 2<br />
CH 2·O·P·O·CH 2<br />
CH·O·CO·R 1<br />
CH 2·O·CO·R 2<br />
Phosphatidylcholine (lecithin) Diphosphatidylglycerol (cardiolipin)<br />
H 2C·O·CO·R 1<br />
R 2·CO·O·CH O<br />
H 2CO·P·O·CH 2·CH 2·NH 2<br />
O –<br />
H 2C·O·CO·R 1<br />
R 2·CO·O·CH O<br />
O –<br />
H·C·O·CO·R 3<br />
H 2·C·O·P·O·CH 2·CH·COOH<br />
Phosphatidylethanolamine (cephalin) Phosphatidylserine (cephalin)<br />
OH OH<br />
H<br />
H<br />
OH<br />
HO<br />
H<br />
H<br />
H OH<br />
O<br />
O P O CH 2<br />
O –<br />
(4.115)<br />
(4.117)<br />
CH·O·CO·R 1<br />
CH 2·O·CO·R 2<br />
O –<br />
(4.122)<br />
NH 2<br />
(4.119)<br />
(4.121)<br />
Fig. 4.31 Examples of polar phospholipids <strong>and</strong> glycolipids (see also 4.23), phosphate residue.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.20 Analyses of the lipid classes in two barley malts (Anness, 1984). The results are<br />
expressed as the fatty acids in each fraction per unit dry weight, (mg fatty acids/g dry weight). This<br />
type of analysis does not reveal the presence of lipids that do not contain fatty acids (e.g. free<br />
sterols)<br />
Class of lipid Weeah malt Sonja malt<br />
PHOSPHOLIPIDS (4.116±4.122)<br />
GLYCOLIPIDS<br />
2.4 3.2<br />
Digalactosyl monoglyceride 0.9 1.1<br />
Digalactosyl diglyceride (4.115) 1.0 1.1<br />
Monogalactosyl monoglyceride 0.2 0.1<br />
Monogalactosyl diglyceride (4.114) 0.4 0.7<br />
NEUTRAL LIPIDS<br />
Acylsterylglycosides 0 0<br />
Monoglycerides (4.103) 0.3 0.3<br />
Free fatty acids (4.107±4.113) 2.0 2.2<br />
Diglycerides (4.104, 4.105) 1.0 1.2<br />
Triglycerides (4.102) 19.8 25.3<br />
Steryl esters 0.3 0.7<br />
Table 4.21 Examples of the major total fatty acids of the lipids found in the grist, the spent grains<br />
<strong>and</strong> the sweet wort from (A) a mash prepared in a mash tun <strong>and</strong> (B) a mash separated in a lauter tun<br />
(Anness <strong>and</strong> Reed, 1985)<br />
Sample Fatty acid composition (%)<br />
16:0 18:0 18:1 18:2 18:3<br />
(4.108) (4.110) (4.111) (4.112) (4.113)<br />
(A)<br />
Grist 21.2 0.8 9.7 58.4 9.9<br />
Spent grains 24.8 1.2 10.1 54.5 9.3<br />
Sweet wort<br />
(B)<br />
47.6 5.2 5.9 36.3 4.9<br />
Grist 20.8 1.0 11.3 57.9 8.9<br />
Spent grains 25.2 0.8 10.5 57.5 6.0<br />
Sweet wort 41.4 3.4 5.4 45.3 4.4<br />
lipid levels in beers are normally so low that their effects on foam are negligible. The<br />
majority view is that the lipids are important, <strong>and</strong> that it is the lipids in the last runnings<br />
from the mash that are responsible for their ability to reduce foam stability.<br />
Greatinterestattachestotheeffectsoflipidsonflavour.Theconcentrationsofthefree<br />
fattyacidsinbeerappeartobetoo lowtohavedirecteffects.However,differentlevels of<br />
free fatty acids can influence the production ofesters by yeast during fermentation<strong>and</strong> so<br />
alter the flavour of the beer produced. The greatest interest is in the effects of lipids on<br />
the flavour deterioration of beer during storage. During mashing some lipid seems to<br />
disappear because it is oxidized, by oxygen dissolved in the mash, to more polar<br />
substances, some of which reach the beer <strong>and</strong>, during storage, give rise to unsaturated<br />
aldehydes (such as trans-2 nonenal <strong>and</strong> trans-2, cis-6-nonadienal) which give the beer an<br />
unpleasant, cardboard like flavour. The chain of reactions is complicated (Fig. 4.32).<br />
Lipids are hydrolysed by lipases (lipid hydrolases) <strong>and</strong> esterases to free fatty acids, a<br />
major proportion of which is linoleic (4.112) <strong>and</strong> linolenic (4.113) acids, which are<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glycerides, glycolipids, phospholipids<br />
H2O Lipases, hydrolases<br />
Many products, including<br />
OH<br />
Linoleic acid (18:2)<br />
Linolenic acid (18:3)<br />
(4.112)<br />
O2 Oxidations<br />
Lipoxidases, LOX<br />
(4.113)<br />
OOH<br />
OOH<br />
Aldehydes<br />
Aldehyde acids<br />
(4.123)<br />
O<br />
OH<br />
O<br />
OH<br />
OOH<br />
OOH<br />
Hydroperoxydi- <strong>and</strong> triene acids (various isomers)<br />
Chain cleavage<br />
Hydroperoxide<br />
lyase<br />
Isomerase Cyclase Isomerase<br />
α- <strong>and</strong> γ-Ketol<br />
fatty acids<br />
Cyclic<br />
acids<br />
unsaturated. Some of these acids may have been oxidized while still combined in the<br />
original lipid. Malt acrospires are rich in lipases <strong>and</strong> lipid degrading enzymes. Lipases are<br />
active to some extent during mashing. The unsaturated acids are partly oxidized by<br />
oxygen in the presence of lipoxidase enzymes (LOX, two isoenzymes occur), <strong>and</strong> perhaps<br />
peroxidase, giving rise to several unsaturated hydroperoxides (4.123). Some autoxidation<br />
may also occur, but it seems that enzyme-catalysed oxidation is the most important, <strong>and</strong><br />
this may be reduced by reducing the mash pH from 5.5 to 5.0 (Kobayashi et al., 1993).<br />
The diene <strong>and</strong> triene systems of linoleic (4.112) <strong>and</strong> linolenic (4.113) acids are very<br />
readily oxidized.<br />
Under the influence of hydroperoxide isomerases (which are relatively heat stable) <strong>and</strong><br />
other enzymes a complex mixture of substances is formed, including saturated <strong>and</strong><br />
unsaturated aldehydes <strong>and</strong> aldehyde acids, ketols, cyclic compounds, epoxyhydroxyacids<br />
<strong>and</strong> trihydroxyfatty acids, including various trihydroxyoctadecenoic acid isomers (Fig.<br />
4.32). The aldehydes contribute to a cardboard flavour, but most of these substances are<br />
O<br />
OH<br />
Epoxy-hydroxyene<br />
fatty acids<br />
H2O<br />
Hydrase<br />
Trihydroxy unsaturated<br />
fatty acids<br />
Fig. 4.32 Possible stages in the oxidative breakdown of the major unsaturated fatty acids during<br />
mashing (after Briggs, 1978; Gardner, 1988). The number of possible products is very large indeed.<br />
It is thought that the unsaturated trihydroxy-fatty acids are the precursors of staling flavour<br />
compounds in beers.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
O<br />
O<br />
OH<br />
OH
lost by evaporation during the hop-boil <strong>and</strong> the rest are reduced to the corresponding<br />
alcohols by the yeast. These alcohols are not oxidized during beer storage <strong>and</strong> they are<br />
not the source of the `staling aldehydes' in beer. It seems that these arise, by some<br />
unknown mechanism, from the various unsaturated trihydroxyacids which are not<br />
retained in the mash, are easily water soluble <strong>and</strong> which survive into the beer (MoÈller-<br />
Hergt et al., 1999). The trihydroxy fatty acids have foam-collapsing properties.<br />
Increasingly, ways to minimize oxidation during mashing are being sought, to minimize<br />
theformationofprecursorsof`off-flavours'.Minimumquantitiesoffattyacid-containing<br />
lipids are extracted from mashes made at 62 64ëC (143.6 147.2ëF). The amounts<br />
extracted from mashes made at 68ëC (154.4ëF) are roughly double those extracted from<br />
the cooler mashes (Forch <strong>and</strong> Runkel, 1974).<br />
4.4.9 Phenols<br />
Cereal grains contain complex mixtures of phenols <strong>and</strong> polyphenols. Barley grain <strong>and</strong><br />
malts have been most studied, but sorghum phenols have also been studied, especially in<br />
dark-grained, `birdproof' cultivars. Barley <strong>and</strong> sorghum seem to be alone among the<br />
cereals in possessing polymeric flavanols. The phenols in barley vary widely in their<br />
complexity.Compoundssuch astyrosine(4.48),tyramine(4.68)<strong>and</strong>hordenine (4.83)are<br />
present as are arange of phenolic acids. These may be divided into two groups, the<br />
substituted benzoic acids <strong>and</strong> the substituted cinnamic acids (Fig. 4.33; Briggs, 1998;<br />
McMurrough et al., 1984; 1996). Of the benzoic acids vanillic acid (4.127) is the most<br />
abundant in wort, while of the cinnamic acids ferulic acid (4.131) is the most abundant in<br />
malt (Table 4.22). The acids occur free <strong>and</strong> in combination, apparently as esters (as in<br />
chlorogenic acid (4.133) <strong>and</strong> glycosides). Ferulic acid occurs free <strong>and</strong> attached to some<br />
arabinose residues in pentosans <strong>and</strong> it is released into wort during mashing, with an<br />
optimum temperature of about 45 ëC (113 ëF). During boiling or under the influence of<br />
some bacteria <strong>and</strong> wild yeasts some ferulic acid is decarboxylated to yield 4-vinyl<br />
guaiacol (4.134) a substance which in most beers confers an undesirable flavour.<br />
However, in wheat beers the presence of this substance is desirable. In some barleys, <strong>and</strong><br />
their malts, coloured anthocyanin pigments (delphinidin (4.135), cyanidin (4.136) <strong>and</strong><br />
perhaps pelargonidin (4.137)) occur, either free or as glycosides, but these substances<br />
seem to have no significance in brewing.<br />
Table 4.22 The concentrations of the major phenolic acids in an unboiled lager wort (McMurrough<br />
et al., 1984)<br />
Benzoic acid derivatives Concentration (mg/litre)<br />
Gallic acid (4.125) 0.1<br />
Protocatechuic acid (4.126) 0.5<br />
4-Hydroxybenzoic acid (4.124) 0.6<br />
Vanillic acid (4.127) 1.4<br />
Syringic acid (4.128) 0.6<br />
Cinnamic acid derivatives<br />
Caffeic acid (4.130) 0.1<br />
p-Coumaric acid (4.129) 0.6<br />
Ferulic acid (4.131) 1.3<br />
Sinapic acid (4.132) 0.4<br />
TOTAL 5.6 (mg/litre)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(4.124)<br />
(4.125)<br />
(4.126)<br />
(4.127)<br />
(4.128)<br />
Benzoic acid series<br />
HO<br />
R1<br />
R2<br />
COOH<br />
p-Hydroxybenzoic acid<br />
(R1 = R2 = H)<br />
Gallic acid<br />
(R1 = R2 = OH)<br />
Protocatechuic acid<br />
(R1 = OH, R2 = H)<br />
Vanillic acid<br />
(R1 = OCH3, R2 = H)<br />
Syringic acid<br />
(R1 = R2 = OCH3)<br />
HO COOH<br />
Cinnamic acid series<br />
CH == CH·COOH<br />
Of more significance are the colourless flavan-3-ols ((+)-catechin (4.138), (-)epicatechin<br />
(4.139), (+)-gallocatechin (4.140) <strong>and</strong> epigallocatechin (4.141)) <strong>and</strong> related<br />
polymeric materials (Fig. 4.34). These four monomeric substances are not proanthocy<strong>and</strong>ins.<br />
In contrast, polymeric flavan-3-ol materials give rise to anthocyanin pigments<br />
when heated in acidic butanol in air. At first this was thought to be because they were<br />
monomeric flavan-3,4-diols, <strong>and</strong> so they were called leucoanthocyanins. This is not<br />
correct <strong>and</strong> at present they are usually called anthocyanogens by brewers <strong>and</strong><br />
proanthocyanidins by chemists. As analytical methods have improved so the great<br />
complexity of this group of substances has been recognized. In arecent study dimers (7),<br />
trimers (19), tetramers (23), <strong>and</strong> pentamers (7) were recognized (Whittle et al., 1999),<br />
<strong>and</strong> this total, of 56, may well increase as analytical methods are further refined. An<br />
example of aprodelphinidin pentamer (4.145) is shown in Fig. 4.35. In addition to the<br />
HO<br />
HO O·CO·CH == CH OH<br />
OH<br />
(4.133) Chlorogenic acid<br />
(3-Caffeoylquinic acid)<br />
(4.129)<br />
(4.130)<br />
(4.131)<br />
(4.132)<br />
OH<br />
R1<br />
R2<br />
p-Coumaric acid<br />
(R1 = R2 = H)<br />
Caffeic acid<br />
(R1 = OH, R2 = H)<br />
Ferulic acid<br />
(R1 = OCH3, R2 = H)<br />
Sinapic acid<br />
(R1 = R2 = OCH3)<br />
OH<br />
OCH3<br />
CH == CH2<br />
(4.134) Vinyl guaiacol<br />
Fig.4.33 Substitutedbenzoic<strong>and</strong>cinnamicacids<strong>and</strong>somerelatedcompoundsthatoccurinsweet<br />
wort.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(4.138)<br />
(4.140)<br />
(4.142)<br />
(4.143)<br />
(4.144)<br />
HO<br />
HO<br />
HO<br />
OH<br />
OH (+) Catechin<br />
OH<br />
OH<br />
O<br />
OH<br />
OH<br />
OH<br />
O<br />
Cl<br />
O<br />
Barley procyanidin B3<br />
(Catechin – catechin)<br />
Prodelphinidin B3<br />
OH<br />
(+) Gallocatechin<br />
OH<br />
Anthocyanin pigments<br />
R1<br />
OH<br />
OH<br />
OH<br />
R2<br />
(4.139)<br />
(4.141)<br />
(R1 = R2 = OH) Delphinidin chloride (4.135)<br />
(R1 = H, R2 = OH) Cyanidin chloride (4.136)<br />
(R1 = R2 = H) Pelargonidin chloride (4.137)<br />
HO<br />
Sorghum<br />
Procyanidin B1<br />
(epicatechin – catechin)<br />
OH<br />
OH<br />
OH<br />
(gallocatechin – catechin) OH OH<br />
OH<br />
O OH<br />
HO<br />
OH<br />
HO<br />
HO<br />
HO<br />
OH<br />
HO<br />
OH<br />
HO<br />
OH<br />
OH<br />
O<br />
OH<br />
O<br />
OH<br />
O<br />
OH<br />
O<br />
OH<br />
O<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH<br />
(–) Epicatechin<br />
OH<br />
OH<br />
O<br />
OH<br />
OH<br />
Epigallocatechin<br />
OH<br />
O<br />
Fig. 4.34 Flavanols, including some that occur in sweet worts.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH
HO<br />
OH<br />
HO<br />
O<br />
OH<br />
HO<br />
OH<br />
O<br />
OH<br />
HO<br />
OH<br />
OH<br />
OH<br />
OH<br />
O<br />
OH<br />
HO<br />
OH<br />
number <strong>and</strong> identity of the constituent monomers the structures are complicated by<br />
differences in stereochemistry, as illustrated by the dimers procyanidin B 1 (4.142),<br />
procyanidin B3 (4.143) <strong>and</strong> prodelphinidin B3 (4.144). Concentrations of some flavans in<br />
barley malts are catechin (4.138) 25 75 mg/kg; prodelphinidin B3 (4.144), 186 362 mg/<br />
kg; procyanidin B3 (4.143), 130 276 mg/kg <strong>and</strong> 4 trimers, 336 671 mg/kg. (McMurrough<br />
<strong>and</strong> Delcour, 1994).<br />
Malt polyphenols are partly dissolved <strong>and</strong> partly destroyed during mashing with pale<br />
malts chiefly, it is thought, by oxidative reactions mostly catalysed by peroxidase,<br />
perhaps with contributions from catalase <strong>and</strong> polyphenol oxidase. The nature of the<br />
phenolic material in wort is influenced strongly by how strongly the malt has been kilned,<br />
the availability of oxygen during mashing <strong>and</strong> the mashing programme. Mashing or<br />
sparging at elevated temperatures extracts more polyphenol into wort. Polyphenols are<br />
largely destroyed in experimental mashes made with green malt, <strong>and</strong> mash aeration or<br />
additions of hydrogen peroxide are most effective at removing anthocyanogens if the<br />
OH<br />
O<br />
OH<br />
OH<br />
OH<br />
A-prodelphinidin pentamer<br />
(4.145)<br />
Fig. 4.35 The formula (ignoring the stereochemistry) of a pentameric proanthocyanidin<br />
(anthocyanogen) from wort (Whittle et al., 1999). The composition may be represented as<br />
g-g-c-c-c, where g <strong>and</strong> c are gallocatechin (4.140) <strong>and</strong> catechin (4.138) units respectively.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
OH<br />
OH<br />
OH<br />
O<br />
OH<br />
OH<br />
OH<br />
OH<br />
OH
malt hasbeen only lightly kilned. Additions of small amounts offormaldehyde tomashes<br />
reduces proanthocyanidin levelsgreatly,probablybylinkingthemtoproteinsinthemash<br />
(Macey, 1970; Macey et al., 1966). Similar results are obtained by mashing with malts<br />
prepared by steeping or re-steeping barley or washing sorghum in dilute solutions of<br />
formaldehyde (Briggs, 1998). All these treatments were tested or used to reduce<br />
proanthocyanidin levels in beers, giving them greater resistance to haze formation, but<br />
their use has been generally discontinued. During wort run-off the mixture of phenolics<br />
altersincomposition<strong>and</strong>thiscanbeofsignificance(Woof<strong>and</strong>Pierce,1966).IfthepHof<br />
the sparge is allowed to increase more phenols are extracted into the wort, which is<br />
undesirable.<br />
When phenols are oxidized they polymerize <strong>and</strong> give rise to red-brown coloured<br />
substances, called phlobaphenes. Some phenols have tanning properties (that is they bind<br />
to proteins <strong>and</strong> may precipitate them), <strong>and</strong> all of them are potentially involved in beer<br />
flavour. However, reports on their significance conflict. Phenolics are credited with<br />
altering the astringency, the mouth-feel <strong>and</strong> the after-taste of beers. However, the<br />
proanthocyanidins are the major `tannins' in malt <strong>and</strong> it is believed that tannins are<br />
responsible for astringency, yet beers made with malts from mutant barleys <strong>and</strong> hop<br />
extracts, both lacking proanthocyanidins, are said to have the same flavour as beers made<br />
from `conventional' barleys. Also the removal of proanthocyanidins from beer by<br />
adsorption on to PVPP (insoluble polyvinyl polypyrrolidone) is reported to have little<br />
effect on flavour. Some phenols have antioxidant effects, that is, they block the oxidative<br />
reactions of other substances, at least sometimes by destroying free radical intermediates.<br />
The significance of malt phenol's antioxidant properties in brewing is unclear. Many<br />
oftheo-diphenolsshouldreadilyoxidizetoo-quinones(Briggs,1998),whichwillreadily<br />
reactwithmanyothersubstances.Theinteractionsduringbrewingbetweenphenolics<strong>and</strong><br />
proteinsarecertainlyimportant.Maltsmadefrommanybirdproofsorghumsaresorichin<br />
phenolic tannins that during mashing many of the enzymes are inhibited <strong>and</strong> conversion<br />
is insufficient. In mashes, <strong>and</strong> indeed in wort, there is apartition of proanthocyanidins<br />
between being free in solution, being bound to soluble proteins <strong>and</strong> being bound to<br />
insoluble proteins. The associations may be reversible, as in chill hazes, or may be<br />
irreversible, as in permanent hazes. More phenolics are extracted from well-modified<br />
malts during mashing because, it is supposed, less insoluble protein remains in the grist<br />
<strong>and</strong>soasmallerproportionofphenolicsarebound.Hopsalsocontributepolyphenolicsto<br />
wort.<br />
4.4.10 Miscellaneous acids<br />
Worts contain awide range of aliphatic acids, in addition to fatty acids (Fig. 4.30), mostly<br />
in small amounts. These include several saturated <strong>and</strong> unsaturated, low molecular weight<br />
fatty acids (C6 C10), mesaconic/laevulinic acid, pyruvic acid (4.146) -ketoglutaric acid<br />
(4.147), fumaric acid (4.148), succinic acid (4.149), lactic acid (4.150), oxalic acid (4.151),<br />
malic acid(4.152), citric acid (4.153), kojic acid (4.154)<strong>and</strong> gluconicacid (4.155).Most of<br />
these are well-known intermediary metabolites (Fig. 4.36; MacWilliam, 1968; Moll, 1991).<br />
The significance of these substances is usually unclear, although they must contribute to the<br />
pH buffering capacity of the wort <strong>and</strong> most can probably be metabolized by yeast.<br />
Two of these materials are of particular interest. Lactic acid ((4.150) a mixture of the<br />
D- <strong>and</strong> L-isomers) arises from the malt <strong>and</strong> the microbes on its surface. Under some<br />
conditions excessive production results in malts that give too acid mashes. Sometimes<br />
lactic acid is deliberately added to mashes (as acidified wort, or as acid malt) to reduce<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH 3·CO·COOH<br />
Pyruvic acid<br />
(4.146)<br />
HOOC·(CH 2) 2·CO·COOH<br />
α-Ketoglutaric acid<br />
(4.147)<br />
HOOC H<br />
C<br />
C<br />
H COOH<br />
Fumaric acid<br />
(4.148)<br />
CH 2·COOH<br />
CH 2·COOH<br />
Succinic acid<br />
(4.149)<br />
CH 3·CH(OH)·COOH<br />
Lactic acid<br />
(4.150)<br />
COOH<br />
COOH<br />
Oxalic acid<br />
(4.151)<br />
HO<br />
HO·CH·COOH<br />
CH 2·COOH<br />
Malic acid<br />
(4.152)<br />
CH 2·COOH<br />
HO·C·COOH<br />
CH 2·COOH<br />
Citric acid<br />
(4.153)<br />
O<br />
O<br />
Kojic acid<br />
(4.154)<br />
COOH<br />
H·C·OH<br />
HO·C·H<br />
H·C·OH<br />
H·C·OH<br />
CH 2OH<br />
(4.155)<br />
CH 2OH<br />
D-Gluconic acid<br />
Fig. 4.36 Formulae of some of the organic acids present in worts.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 4.23 The concentrations of some inorganic ions in various sweet worts (A, MaÈndl, 1974; B,<br />
Rudin, 1974; C, Moll, 1991).The all-malt worts in column A were adjusted to a concentration of<br />
12%. The worts in column B were made with <strong>and</strong> without adjuncts<br />
Ionic species Concentrations (mg/litre)<br />
A B C<br />
Potassium, K +<br />
522 310±770 300±700<br />
Sodium, Na +<br />
Calcium, Ca<br />
25 11±112 10±100<br />
2+<br />
Magnesium, Mg<br />
39 40±62 40±100<br />
2+<br />
130 ± 100±150<br />
Copper, Cu 2+<br />
Iron, Fe<br />
0.08 0.12±0.13 0.1±0.2<br />
3+ (ferric)<br />
Manganese, Mn<br />
0.08 0.23±0.37 0.1±0.3<br />
2+ (manganous) 0.13 ± 0.1±0.2<br />
Zinc, Zn 2+<br />
0.12 0.07±0.16 0.1±0.3<br />
the mash pH in acontrolled fashion. Worts also contain oxalic acid (4.151). This, the<br />
simplest dicarboxylic acid, has astrong affinity for calcium ions <strong>and</strong> readily forms<br />
crystalline precipitates of calcium oxalate. If formed late in the brewing process this can<br />
give rise to oxalate haze <strong>and</strong> the crystals can form nuclei for the release of carbon<br />
dioxide, so potentiating gushing (over-foaming) in beer. During fermentation calcium<br />
oxalate can be deposited on the walls of the fermentation vessel as beer stone. Malted<br />
barleyscancontain5.6 22.8mgoxalicacid/100gdrymatterwhileformaltedwheatsthe<br />
values may be 22.1 50.3mg/100g(Narziss et al., 1986). Presumably much of this is<br />
precipitated in the mash when the liquor contains asufficient concentration of calcium<br />
ions. Since oxalate is potentially toxic it is desirable to reduce its concentration to alow<br />
level.<br />
4.4.11 Inorganic ions in sweet wort<br />
Some 1.5 2.0% of wort solids are materials which, after evaporation <strong>and</strong> combustion,<br />
remain in the ash. Of these the inorganic ions originate in the brewing liquor (Chapter 3)<br />
<strong>and</strong>inthe grist materials,both the malt (around 2 3%ash) <strong>and</strong>the adjuncts(1 3%ash).<br />
Small amounts of ions (copper, iron, nickel, tin, zinc, etc.) may be picked up from the<br />
brewing plant. Later contributions may come from the copper adjuncts <strong>and</strong> hops. The<br />
proportions of the most important ions vary (Table 4.23). Sulphur may be present as<br />
sulphate (200 400mg/litre), in amounts that vary widely with the nature of the brewing<br />
liquor. Sulphuralsooccursinorganic combinationsintheamino acidsmethionine(4.41),<br />
cysteine (4.31) <strong>and</strong> cystine (4.32), the tri-peptide glutathione ( -glutamyl-cysteinylglycine),<br />
coenzyme A, mercaptans, as well as hydrogen sulphide <strong>and</strong> sulphite ions.<br />
Chloride ions (40±500mg/litre) <strong>and</strong> phosphate ions (500±900mg/litre) also occur<br />
(Briggs, 1998; Lee, 1990; MacWilliam, 1968; Moll, 1991).<br />
Little phosphate comes from the brewing liquor, except where phosphoric acid or acid<br />
phosphate salts have been used for pH adjustment. Most comes from the grist <strong>and</strong> may<br />
originate from nucleic acids, nucleotides, phospholipids (Fig. 4.31), <strong>and</strong> especially myoinositol<br />
hexaphosphate (phytic acid, 4.156), of which there may be 0.6±0.9% in the dry<br />
malt (Lee, 1990). Phytic acid is a strong chelating agent, <strong>and</strong> binds copper, iron <strong>and</strong> zinc<br />
ions as well as calcium ions. This material undergoes some hydrolysis during mashing,<br />
inorganic phosphate being removed sequentially from successively lower phosphate<br />
esters until free myo-inositol (4.91) is released. The extent of phytate hydrolysis is<br />
dependent on the level of the enzyme (or enzymes) with phytase activity remaining in the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
malt after kilning <strong>and</strong> the temperature programme used in mashing. Lightly kilned malts<br />
may contain about aquarter of the enzyme originally present, while ale malts may<br />
contain little active enzyme. The pH optimum of the enzyme is about 5, <strong>and</strong> the optimum<br />
temperatureis45 50ëC(113 122ëF).Theinteractionsbetweenphosphates<strong>and</strong>calcium<br />
ions contribute to the desirable fall in pH which occurs during mashing (Chapter 3).<br />
Usually about two-thirds of the phosphate in sweet wort is inorganic phosphate.<br />
It seems that usually wort provides all the necessary inorganic ions that yeast requires<br />
for sound fermentations, with the occasional exception of zinc (Donhauser, 1986; Jacobsen,<br />
1986; Lie et al., 1980). Because not all the ash components are extracted during mashing<br />
the ionic composition of the sweet wort is not the simple sum of the ions initially present in<br />
the grist <strong>and</strong> the mashing liquor (Table 4.24). Less than 5% of the zinc (or copper or iron)<br />
present in the grist is dissolved during mashing, <strong>and</strong> the proportions dissolved can be very<br />
variable. Of the zinc present in the wort only a proportion is available to yeast, presumably<br />
because the remainder is chelated or otherwise bound to other substances. Consequently<br />
simple analyses of wort zinc contents are unreliable for predicting zinc deficiency.<br />
Probably a 12% wort should contain at least 0.08 <strong>and</strong> preferably 0.1 0.2 mg Zn/litre to<br />
ensure a good fermentation. Where permitted, traces of a soluble zinc salt may be added to<br />
the wort. Where this is not permitted the use of well-modified malts <strong>and</strong> carefully acidified<br />
mashes reduce problems of zinc deficiency, as does the use of mashing equipment with<br />
metal components from which traces of zinc can dissolve.<br />
4.5 Mashing <strong>and</strong> beer flavour<br />
H<br />
O<br />
H2O3P PO 3H 2 PO 3H 2<br />
O O PO 2H 2<br />
H H<br />
OPO2H3 H<br />
H<br />
Phytic acid, myo-inositol hexaphosphate<br />
(4.156)<br />
Malts <strong>and</strong> other components of the grist influence beer flavour. However, the<br />
relationships are extremely complicated. During mashing flavoured substances are<br />
extracted into the wort. Some will be destroyed or partly or wholly lost during the hopboil,<br />
while other `flavour precursors' will be converted into flavoursome substances, <strong>and</strong><br />
others will reach the fermenter unchanged. The yeast may then metabolize many of these<br />
O<br />
H<br />
O<br />
PO3H2 Table 4.24 The ionic compositions (mg/litre) of a brewing water, a sweet wort made with it <strong>and</strong><br />
the beer (Rudin, 1974)<br />
Liquid Ca 2+<br />
Mg 2+<br />
Na +<br />
K +<br />
Cl SO 2 4 PO 3 4 HCO 2 3<br />
<strong>Brewing</strong> liquor 169 36 55 6 247 205 ± 165<br />
Wort (12 ëPlato) 165 127 101 550 450 338 846 ±<br />
Beer (12 ëPlato,<br />
original gravity) 168 113 110 440 420 330 520 ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
substances altering their flavours. Some malt flavour substances, such as vanillin <strong>and</strong> -<br />
phenyl ethanol, partly occur combined as -glycosides. During mashing they can be<br />
released by hydrolysis, catalysed by -glucosidase. In addition, oxidative changes<br />
occurring to the lipids during mashing can give rise to precursors of staling flavours. The<br />
`redox state' of the beer has an influence on the rate of flavour deterioration <strong>and</strong> haze<br />
formation (Bamforth et al., 1993, 1997; Briggs, 1998; Moir, 1989; Van Den Berg <strong>and</strong><br />
Van Eerde, 1982). It can be advantageous to exclude oxygen from the mash <strong>and</strong><br />
subsequently, during the production of hopped wort.<br />
Higher levels of `anti-oxidant' substances in the beer retard deterioration processes.<br />
These agents, which originate in the grist, seem to work in different ways. They compete<br />
for oxygen by being oxidized themselves, or they inhibit enzymes catalysing oxidations,<br />
<strong>and</strong>/or they `scavenge' free radicals. Anti-oxidants include sulphite <strong>and</strong> bisulphite ions,<br />
polyphenols <strong>and</strong> reductones, which are ene-diol substances resembling ascorbic acid<br />
(4.96), which are formed during Maillard reactions. These compounds are found in dark<br />
malts, which have long been known to have flavour-stabilizing properties. Many<br />
hundreds of potentially active flavour substances are derived from malts or adjuncts <strong>and</strong><br />
include aldehydes, ketones, amines, thiols <strong>and</strong> other sulphur-containing substances,<br />
heterocyclic oxygen-, nitrogen- <strong>and</strong> sulphur-containing substances <strong>and</strong> phenols. Sparging<br />
may extract unwanted flavours. Moderating sparge temperatures, for example up to 75ëC<br />
(167ëF), <strong>and</strong> keeping the pH low, e.g., to 5.5, improves beer flavour <strong>and</strong> stability.<br />
Asubstance that has attracted particular interest is dimethyl sulphide (DMS, (4.158)<br />
Fig.4.37;Dickenson,1983).InsomeEuropeanlagersappreciablelevelsofthissubstance<br />
are desirable while in some other beers its absence is preferred. The precursor of this<br />
highly volatile material is S-methyl methionine (SMM, 4.157), asulphonium compound<br />
formed by the metabolic methylation of methionine (4.41) in the malt. This substance is<br />
heatlabile<strong>and</strong>sowillonlysurviveinmaltifthisislightlykilned.Someisdecomposedto<br />
DMS <strong>and</strong> homoserine (4.159) <strong>and</strong> the DMS produced is mostly lost with the kilning air.<br />
Some is oxidized to the less volatile dimethyl sulphoxide (DMSO, 4.160). More SMM is<br />
decomposed during the hop-boil. Surviving DMS, SMM <strong>and</strong> small amounts of DMSO<br />
reach the fermenter. Yeast may reduce the DMSO to DMS. Thus the level of DMS<br />
present in abeer depends on the malt used <strong>and</strong> the details of the production process.<br />
CH 3<br />
CH 3<br />
S·CH 2·CH 2·CH·COOH<br />
NH 2<br />
+ OH<br />
CH 3<br />
CH 3<br />
S-Methyl methionine (SMM) (4.158) Dimethyl<br />
(4.157)<br />
sulphide (DMS)<br />
Oxidation Reduction<br />
CH 3<br />
CH 3<br />
S + HO·CH 2·CH 2·CH·COOH<br />
S==O<br />
Dimethyl sulphoxide (DMSO)<br />
(4.160)<br />
NH 2<br />
Homoserine<br />
(4.159)<br />
Fig. 4.37 The formation of dimethylsulphide ((4.122), DMS) from S-methyl methionine ((4.121),<br />
SMM, the DMS precursor, DMS-P) <strong>and</strong> the interconversions of DMS <strong>and</strong> dimethylsulphoxide<br />
((4.123), DMSO).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
4.6 Spent grains<br />
At the end of every mash abrewer has aload of spent grains or draff, which must be<br />
disposed of. Of the original grist some 17 22% of the original dry matter remains,<br />
about 18 20kg fresh weight/hl beer produced. This material is wet <strong>and</strong>, depending on<br />
the mashing system employed, contains up to 80% water. Liquid from this wet<br />
material is turbid <strong>and</strong> represents an effluent that, because of its high BOD, is costly to<br />
dispose of if directed to drain (Chapter 3). Obtaining representative samples of the<br />
spent grains is difficult, because of inhomogeneities in the filter bed, but they should<br />
be investigated to evaluate the mashing process. It is desirable to determine their<br />
content of residual extract (which is inversely related to extract recovery in the wort),<br />
<strong>and</strong> residual starch which, ideally, would have been converted during mashing. In<br />
addition inspection of the solids shows how well milling has been carried out <strong>and</strong> to<br />
what extent husk material has remained intact. The wet grains may be squeezed to<br />
remove some liquor (minimizing drainage <strong>and</strong> h<strong>and</strong>ling problems) <strong>and</strong> the squeezed<br />
draff may be dried with hot air. Drying is unusual as it is costly (Chapter 3). Often the<br />
squeeze liquor is returned, with or without treatment, to the process stream to recover<br />
the extract.<br />
The composition of the spent grains depends on the grist used, the mashing<br />
programme <strong>and</strong> the recovery of the extract (Table 4.25). Relative to the malt, spent grains<br />
are enriched in protein, fibre, ash <strong>and</strong> lipid. Their major use is as a valuable feed for<br />
ruminants. They may be fed directly or after ensilage. As they deteriorate rapidly draff<br />
should be removed from the brewery as quickly as possible, to prevent their becoming a<br />
source of undesirable organisms. It has been suggested that they should be treated with a<br />
preservative, such as propionic acid, to minimize decomposition. Other actual or<br />
proposed uses for spent grains are inclusion in mushroom compost, <strong>and</strong> use as a substrate<br />
for the cultivation micro-organisms, for example filamentous fungi for feeding pigs. The<br />
draff might be fed to the pigs directly, after the mechanical removal of the fibre. After<br />
composting the spent grains might be used for turf management or as a horticultural soil<br />
conditioner. It has also been proposed that the grains be digested with enzyme<br />
preparations to produce extra `wort' to use in the brewery or as a culture medium for<br />
other microbes.<br />
Table 4.25 The composition <strong>and</strong> nutritive value (for sheep) ranges of brewers' wet spent grains<br />
(Briggs et al., 1981). The values in parentheses are the digestible amounts. Metabolizable energy ˆ<br />
digestible energy 0.81<br />
Mean Range<br />
Dry matter (%) 26.3 24.4±30.0<br />
Crude protein (%) 23.4 (18.5) 18.4±26.2 (13.9±21.3)<br />
Crude fibre (%) 17.6 (7.9) 15.5±20.4 (6.6±10.2)<br />
Ether extract (%) 7.7 (7.7) 6.1±9.9 (5.6±9.2)<br />
Total ash (%) 4.1 3.6±4.5<br />
Digestible energy (mJ/kg dry wt.) 13.8 13.0±14.8<br />
Gross energy (mJ/kg dry wt.) 21.4 21.1±21.8<br />
DOMD* in vivo (%) 59.4 55.2±64.3<br />
DOMD* in vitro (%) 48.6 44.8±51.5<br />
* DOMD ˆ Digestibility of organic matter (dry).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
4.7 References<br />
ALBINI, P. A., BRIGGS, D. E. <strong>and</strong> WADESON, A. (1987) J. Inst. <strong>Brewing</strong>, 93, 97.<br />
ALLEN, W. G. (1987) Brewer's Digest, 62 (3), 18.<br />
ANDERSON, I. W. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, 193.<br />
ANNESS, B. J. (1984) J. Inst. <strong>Brewing</strong>, 90, 315.<br />
ANNESS, B. J. <strong>and</strong> REED, R. J. R. (1985) J. Inst. <strong>Brewing</strong>, 91, 82, 313.<br />
BAJOMO, M. F. <strong>and</strong> YOUNG, T. W. (1993) J. Inst. <strong>Brewing</strong>, 99, 153.<br />
BAMFORTH, C. W., MULLER, R. E. <strong>and</strong> WALKER, M. D. (1993) J. Amer. Soc. Brew. Chem., 51, 79.<br />
BAMFORTH, C. W., MOORE, J., MCKILLOP, D., WILLIAMSON, G. <strong>and</strong> KROON, P. A. (1997) Proc. 26th Congr.<br />
Eur. Brew. Conv., Maastricht, 75.<br />
BARRETT, J., CLAPPERTON, J. F., DIVERS, D. M. <strong>and</strong> RENNIE, H. (1973) J. Inst. <strong>Brewing</strong>, 79, 407.<br />
BARRETT, J., BATHGATE, G. N. <strong>and</strong> CLAPPERTON, J. F. (1975) J. Inst. <strong>Brewing</strong>, 81, 31.<br />
BATHGATE, G. N. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, 198.<br />
BATHGATE, G. N. <strong>and</strong> DALGLEISH, C. E. (1975) Proc. Amer. Soc. Brew. Chem., 33, 32.<br />
BATHGATE, G. N. <strong>and</strong> PALMER, G. H. (1972) Die StaÈrke, 24 (10), 336.<br />
BOURNE, D. T. (1998) Proc. 5th Aviemore Conf. Malting, <strong>Brewing</strong> <strong>and</strong> Distilling. (Campbell, I. ed.).<br />
London, The Inst. of <strong>Brewing</strong>, p. 237.<br />
BOURNE, D. T. <strong>and</strong> WHEELER, R. E. (1982) J. Inst. <strong>Brewing</strong>, 88, 324.<br />
BOURNE, D. T. <strong>and</strong> WHEELER, R. E. (1984) J. Inst. <strong>Brewing</strong>, 90, 306.<br />
BOURNE, D. T., POWLESLAND, T. <strong>and</strong> WHEELER, R. E. (1982) J. Inst. <strong>Brewing</strong>, 88, 371.<br />
BRENNER, M. W. (1972) M. M. B. A. Tech. Quart., 9, 12.<br />
BRIGGS, D. E. (1978) Barley. London. Chapman <strong>and</strong> Hall, 612 pp.<br />
BRIGGS, D. E. (1992) in Barley: genetics, biochemistry, molecular biology, <strong>and</strong> biotechnology. (Shewry, P.<br />
R. ed.) p. 361. Wallingford. C.A.B. International.<br />
BRIGGS, D. E. (1998) Malts <strong>and</strong> Malting. London. Blackie Academic, 796 pp.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting <strong>and</strong> <strong>Brewing</strong> Science (2nd edn),<br />
Vol. 1. Malt & Sweet Wort. London, Chapman <strong>and</strong> Hall, 387 pp.<br />
BRYCE, J. H., KIM, N. J., ZAMMIT, R., MCCAFFERTY, C. A., PINHEIRO, M. G. M. <strong>and</strong> KENDALL, G. L. (1995)<br />
Proc. 25th Congr. Eur. Brew. Conv., Brussels, p. 285.<br />
BRYCE, J. H., COOPER, D. <strong>and</strong> STEWART, G. G. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht,<br />
p. 357.<br />
BUÈ HLER, T. N., MCKECHNIE, M. F. <strong>and</strong> WAKEMAN, R. J. (1996) Food <strong>and</strong> Bioproducts Processing, 74 (no.<br />
C4), 207.<br />
BURGER, W. C. <strong>and</strong> SCHROEDER, R. L. (1976) J. Amer. Soc. Brew. Chem., 34, 133; 138.<br />
BUTTIMER, E. T. <strong>and</strong> BRIGGS, D. E. (2000) J. Inst. <strong>Brewing</strong>, 106, 71, 83.<br />
BUTTON, A. H. <strong>and</strong> PALMER, J. R. (1974) J. Inst. <strong>Brewing</strong>, 80, 206.<br />
COMRIE, A. A. D. (1967) J. Inst. <strong>Brewing</strong> 74, 335.<br />
COULTATE, T. P. (2002) Food: the chemistry of its components. Cambridge. The Royal Society of<br />
Chemistry, 432 pp.<br />
CRABB, D. <strong>and</strong> BATHGATE, G. N. (1973) J. Inst. <strong>Brewing</strong>, 79, 519.<br />
DEBYSER, W., DERDELINCKX, G. <strong>and</strong> DELCOUR, J. A. (1997a) J. Cereal Sci., 26, 67.<br />
DEBYSER, W., DERDELINCKX, G. <strong>and</strong> DELCOUR, J. A. (1997b) J. Amer. Soc. Brew. Chem., 55, 153.<br />
DE CLERCK (1957) A Textbook of <strong>Brewing</strong>, Vol. 1. (Transl. Barton-wright, K.) London, Chapman <strong>and</strong><br />
Hall, p. 587.<br />
DEMUYAKAR, B., OHTA, Y., NAKATANI, K., FUKUI, N. <strong>and</strong> KANAGAWA, K. (1994) J. Amer. Soc. Brew.<br />
Chem., 52, 111.<br />
DICKENSON, C. J. (1983) J. Inst. <strong>Brewing</strong>, 89, 41.<br />
DIXON, M. <strong>and</strong> WEBB, E. (1958) Enzymes. London. Longmans, Green & Co., 784 pp.<br />
DONHAUSER, S. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C. ± Symposium on Wort Production,<br />
Maffliers, p. 176.<br />
ELIASSON, A.-C. <strong>and</strong> TATHAM, A. (2001) in Cereals <strong>and</strong> Cereal Products ± Chemistry <strong>and</strong> Technology.<br />
(Dendy, D. A. V. <strong>and</strong> Dobraszczyk, B. J. eds) Gaithersburg. Aspen Publishers, p. 68.<br />
ENARI, T. M. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 73.<br />
ENARI, T. M. (1986) Cerevisia, 11 (1), 19.<br />
ENARI, T. M., MIKOLA, J. <strong>and</strong> LINKO, M. (1964) J. Inst. <strong>Brewing</strong>, 70, 405.<br />
ENEVOLDSEN, B. S. (1974) Eur. Brew. Conv. Monograph-I. E.B.C.-Wortsymposium, Zeist, p. 158.<br />
ENEVOLDSEN, B. S. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, p. 683.<br />
ENGAN, S. (1981) <strong>Brewing</strong> Science, Vol. 2 (Pollock, J. R. A. ed.). London, Academic Press, p. 93.<br />
EVANS, E., MA, Y., EGLINTON, J., LANGRIDGE, P., LOGUE, S. <strong>and</strong> BARR, A. (2002) Proc. Congr. Inst. Brew.<br />
(Asia Pacific Sect.)., Adelaide, CD No. 12.<br />
FINCHER, G. B. (1992) in Barley: Genetics, Biochemistry, Molecular Biology <strong>and</strong> Biotechnology (Shewry,<br />
P. R., ed.). Wallingford. C. A. B.-International, p. 413.<br />
FORAGE, A. J. <strong>and</strong> LETTERS, R. (1986) Proc. 19th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Sect.),<br />
Hobart, p. 79.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
FORCH, M. <strong>and</strong> RUNKEL, U.-D. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist,<br />
p. 269.<br />
FORREST, I. S., DICKSON, J. E. <strong>and</strong> SEATON, J. C. (1985) Proc. 20th Congr. Eur. Brew.Conv., Helsinki,<br />
p. 363.<br />
GARDNER, H. W. (1988) in Advanc. Cereal Sci. <strong>and</strong> Technol. IX (Pomeranz, Y. ed.) St. Paul. A. A. C. C.<br />
p. 161.<br />
GEIGER, K. (1972) M. B. A. A. Tech. Quart. 9, 195.<br />
GJERTSEN, P. <strong>and</strong> HARTLEV, P. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the<br />
relationship between malt <strong>and</strong> beer, Helsinki, p. 186.<br />
HALFORD, M. H. <strong>and</strong> BLAKE, G. G. (1972) Proc. 12th Conf. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong><br />
Sect.), Perth, p. 85.<br />
HALL, R. D., HARRIS, G. <strong>and</strong> MACWILLIAM, I. C. (1956) J. Inst. <strong>Brewing</strong>, 62, 232.<br />
HALLGREN, L. (1995) in Sorghum <strong>and</strong> Millets; Chemistry <strong>and</strong> Technology. (Dendy, D. A. V. ed.). St. Paul.<br />
A. A. C. C., p. 283.<br />
HAN, J. Y. <strong>and</strong> SCHWARZ, P. B. (1996) J. Amer. Soc. Brew. Chem., 54, 216.<br />
HARRIS, G. <strong>and</strong> MACWILLIAM, I. C. (1961) J. Inst. <strong>Brewing</strong>, 67, 144, 154.<br />
HIND, H. L. (1950) <strong>Brewing</strong> Science <strong>and</strong> Practice Vol II. <strong>Brewing</strong> Processes. London. Chapman <strong>and</strong> Hall,<br />
pp. 507±1020.<br />
HOME, S., PIETILAÈ , K. <strong>and</strong> SJOÈ HOLM, K. (1993) J. Amer. Soc. Brew. Chem., 51, 108.<br />
HOME, S., LAUREN, A. <strong>and</strong> AUTIO, K. (1994) Proc. 23rd Conf. Inst. <strong>Brewing</strong>, (Asia Pacific Sect.), Sydney,<br />
p. 52.<br />
HOPKINS, R. H. <strong>and</strong> KRAUSE, B. (1947) Biochemistry Applied to Malting <strong>and</strong> <strong>Brewing</strong>. London. Allen &<br />
Unwin, 342 pp.<br />
HUDSON, J. R. (1975) Eur. Brew, Conv. Monograph-II. E. B. C. Barley <strong>and</strong> Malting Symposium, Zeist,<br />
p. 271.<br />
HUG, H. <strong>and</strong> PFENNINGER, H. B. (1979) Proc. 17th Congr. Eur. Brew. Conv., Berlin (West), p. 355.<br />
HUITE, N. J. <strong>and</strong> WESTERMANN, D. H. (1974) Proc. 14th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong><br />
Sect.), Brisbane, p. 133.<br />
HUMBERSTONE, F. J. <strong>and</strong> BRIGGS, D. E. (1998) Proc. 5th Aviemore Conf. Malting, <strong>Brewing</strong> <strong>and</strong> Distilling<br />
(Campbell, I. ed.) London. The Institute of <strong>Brewing</strong>, p. 246.<br />
IMBERTY, A., BULEON, A. <strong>and</strong> VINH TRAN (1991) StaÈrke, 43 (10), 375.<br />
ISHERWOOD, N. D., KIRBY, W., WHEELER, R. E. <strong>and</strong> JONES, M. (1977) Proc. 16th Congr. Eur. Brew. Conv.,<br />
Amsterdam, p. 457.<br />
JACOBSEN, T. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort Production,<br />
Maffliers. p. 196.<br />
JONES, M. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 94.<br />
JONES, M. <strong>and</strong> PIERCE, J. (1967) J. Inst. <strong>Brewing</strong>, 73, 342.<br />
KIENINGER, H. (1969) Proc. 12th Congr. Eur. Brew. Conv., Interlaken, p. 139.<br />
KIENINGER, H. (1972) Brauwelt, 112, 1535; 1793.<br />
KOBAYASHI, N., KANEDA, H., KANO, Y. <strong>and</strong> KOSHINO, S. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo,<br />
p. 405.<br />
KROTTENTHALER, M., ZUÈ RCHER, J., SCHNEIDER, J., BACK, W. <strong>and</strong> WEISSER, H. (1999) Proc. 27th Congr.<br />
Eur. Brew. Conv., Cannes, p. 603.<br />
KUNZE, W. (1996) Technology <strong>Brewing</strong> <strong>and</strong> Malting. (Wainwright, T. transl.). Berlin. VLB, pp. 726.<br />
LAING, H. <strong>and</strong> TAYLOR, D. G. (1984) Proc. 18th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Sect.),<br />
Adelaide, p. 109.<br />
LEE, W. J. (1990) J. Amer. Soc. Brew. Chem., 48, 62.<br />
LETTERS, R. (1992) Ferment, 5, 268.<br />
LETTERS, R. (1994) Proc. 4th Aviemore Conf. Malting, <strong>Brewing</strong> <strong>and</strong> Distilling. (Campbell, I. <strong>and</strong> Priest,<br />
F. G. eds). London, the Inst. of <strong>Brewing</strong>, p. 128.<br />
LETTERS, R. (1995a) Proc. 5th Sci. Tech. Conv. Inst. <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.),<br />
Victoria Falls, p. 115.<br />
LETTERS, R. (1995b) Ferment, 8, 301.<br />
LETTERS, R., HURLEY, J. C. <strong>and</strong> HORAN, H. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.Symposium on<br />
Wort Production, Maffliers, p. 250.<br />
LEWIS, M. J. <strong>and</strong> OH, S. S. (1985) M.B.A.A. Tech. Quart., 22, 108.<br />
LIE, S., HAGE, T. <strong>and</strong> JACOBSEN, T. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C. Symposium on the<br />
relationship between malt <strong>and</strong> beer, Helsinki, p. 213.<br />
LISBJERG, S. <strong>and</strong> NIELSEN, H. (1991) Proc. 3rd Sci. Tech. Conv. Inst. <strong>Brewing</strong> (Central <strong>and</strong> Southern<br />
African Section), Victoria Falls, p. 157.<br />
LITTLE, B. T. (1994) Ferment, 7, 163.<br />
LUCHSINGER, W. W., ENGLISH, H. <strong>and</strong> KNEEN, E. (1958) Proc. Ann. Mtg. Amer. Soc. Brew. Chem., p. 40.<br />
LUÈ ERS, H., KRAUSS, G., HARTMANN, O. <strong>and</strong> VOGT, H. (1934) Wochensch. f. Brau. 51 (46), 361.<br />
MACEY, A. (1970) Proc. 11th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> sect.), Hobart, p. 117.<br />
MACEY, A., STOWELL, K. C. <strong>and</strong> WHITE, H. B. (1966) J. Inst. <strong>Brewing</strong>, 72, 29.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
MACFADDEN, D. P. (1989) Proc 2nd Sci. Tech. Conv. Inst. <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.),<br />
Johannesburg, p. 306.<br />
MACGREGOR, A. W. (1990) Proc. 3rd Aviemore Conf. On Malting, <strong>Brewing</strong> <strong>and</strong> Distilling. (Campbell, I.<br />
ed.). London. Inst. <strong>Brewing</strong>, p. 10.<br />
MACGREGOR, A. W. <strong>and</strong> YIN, X. S. (1990) J. Amer. Soc. Brew. Chem., 48, 82.<br />
MACWILLIAM, I. C. (1968) J. Inst. <strong>Brewing</strong>, 74, 38.<br />
MACWILLIAM, I. C. (1972) J. Inst. <strong>Brewing</strong>, 78, 76.<br />
MACWILLIAM, I. C. (1975) J. Inst. <strong>Brewing</strong>, 81, 65.<br />
MACWILLIAM, I. C., HUDSON, J. R. <strong>and</strong> WHITEAR, A. L. (1963) J. Inst. <strong>Brewing</strong>, 69, 303.<br />
MAÈ NDL, A. (1974) Brauwiss. 27, 177.<br />
MARTIN, P. A. (1978) Brewer's Guard., Aug., p. 29.<br />
MARTIN, P. A. (1979) J. Inst. <strong>Brewing</strong>, 85, 290.<br />
MAULE, D. R. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort Production,<br />
Maffliers, p. 72.<br />
MAULE, D. R. <strong>and</strong> CRABB, D. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the<br />
relationship between malt <strong>and</strong> beer, Helsinki, p. 169.<br />
MCMURROUGH, I. <strong>and</strong> DELCOUR, J. A. (1994) Ferment, 7, 175.<br />
MCMURROUGH, I., ROCHE, G. P. <strong>and</strong> CLEARY, K. G. (1984) J. Inst. <strong>Brewing</strong>, 90, 181.<br />
MCMURROUGH, I., MADIGAN, D., DONNELLY, D., HURLEY, J., DOYLE, H., HENNIGAN, G. <strong>and</strong> MCNULTY, N.<br />
(1996) J. Inst. <strong>Brewing</strong>, 102, 327.<br />
MEDDINGS, P. J. <strong>and</strong> POTTER, O. E. (1971) J. Inst. <strong>Brewing</strong>, 77, 246.<br />
MIKOLA, J., PIETILAÈ , K. <strong>and</strong> ENARI, T.-M. (1972) J. Inst. <strong>Brewing</strong>, 78, 384.<br />
MOIR, M. (1989) Brewer's Guard., Sept., p. 64.<br />
MOLL, M. (1991) Beers <strong>and</strong> Coolers. (Wainwright, T. transl.) Andover. Intercept, p. 495.<br />
MOLL, M., FLAYEUX, R., VINH THAT <strong>and</strong> MARTIN, J. (1974) Eur. Brew. Conv. Monograph-I.<br />
Wortsymposium, Zeist, p. 41.<br />
MOLL, M., FLAYEUX, R., MATHIEU, D. <strong>and</strong> PHAN TAN LUU, R. (1982) J. Inst. <strong>Brewing</strong>, 88, 139.<br />
MOÈ LLER-HERGT, G., WACKERBAUER, K., TRESSL, R., GARBE, L.-A. <strong>and</strong> ZUFULL, C. (1999) Proc. 27th Congr.<br />
Eur. Brew. Conv., Cannes, p. 123.<br />
MORRAYE, C. (1938) Pet. J. Brasseur., 46, 464.<br />
MORRISON, W. R. (1978) in Advanc. Cereal Sci. <strong>and</strong> Technol. II (Pomeranz, Y. ed.) p. 221.<br />
MORRISON, W. R. (1988) J. Cereal Sci., 8, 1.<br />
MULLER, R. F. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 283.<br />
MULLER, R. (1991) J. Inst. <strong>Brewing</strong>, 97, 85, 93.<br />
MULLER, R. (1995) J. Amer. Soc. Brew. Chem., 53, 53.<br />
MUTS, G. C. J. (1991) Proc. 3rd Sci. Tech. Conv. Inst. <strong>Brewing</strong>, (Central & Southern African Section),<br />
Victoria Falls, p. 51.<br />
MUTS, G. C. J. <strong>and</strong> PESMAN, L. (1986) Eur. Brew. Conv. Monograph-XI. E. B. C.-Symposium on Wort<br />
Production, Maffliers, p. 25.<br />
MUTS, G. C. J., KAKEBEEKE, M. G., PESMAN, L. C. <strong>and</strong> VAN DEN BERG, R. (1984) Proc. 18th Conv. Inst.<br />
<strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Section), Adelaide, p. 115.<br />
NARZISS, L. (1977) Brauwelt, 117 (37), 1420.<br />
NARZISS, L. (1978) Proc. 15th Congr. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Sect.), Christchurch,<br />
p. 35.<br />
NARZISS, L. (1991) Brauwelt Internat., 4, 284.<br />
NARZISS, L. (1992a) Die Bierbrauerei, 2 B<strong>and</strong>. Die Technologie der WuÈrzebereitung (7th edn). Stuttgart.<br />
Ferdin<strong>and</strong> Enke, p. 402.<br />
NARZISS, L. (1992b) Brauwelt, (23), 1072.<br />
NARZISS, L., REICHENEDER, E. <strong>and</strong> IWAN, H.-J. (1986) Monatss. f. Brauwiss., 39 (1), 4.<br />
NARZISS, L., MIEDANER, H. <strong>and</strong> NITZCHE, F. (1990) Monatss. f. Brauwiss., 3, 96.<br />
OLIVER-DAUMEN, B. (1988) Brauwelt Internat., (III), 256, 370.<br />
PIERCE, J. S. (1980) Eur. Brew. Conv. Monograph-VI. E. B. C.-Symposium on the relationship between<br />
malt <strong>and</strong> beer, Helsinki, p. 179.<br />
PIERCE, J. S. (1982) J. Inst. <strong>Brewing</strong>, 88, 228.<br />
POLLOCK, J. R. A. <strong>and</strong> POOL, A. A. (1968) Proc. Ann. Mtg. Amer. Soc. Brew. Chem., p. 33.<br />
PRECHTL, C. (1967) M. B. A. A. Tech. Quart., 4, 98.<br />
RENNIE, H. <strong>and</strong> BALL, K. (1979) J. Inst. <strong>Brewing</strong>, 85, 247.<br />
RUDIN, A. D. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 239.<br />
SCHILD, D. (1936) Wochenschr. Brau., 53, 345, 353.<br />
SCHUR, F., PFENNINGER, H. B. <strong>and</strong> NARZISS, L. (1973) Proc. 14th Congr. Eur. Brew, Conv., Salzburg,<br />
p. 149.<br />
SCHUR, F., PFENNINGER, H. B. <strong>and</strong> NARZISS, L. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, p. 191.<br />
SCHWARZ, P. B. <strong>and</strong> HAN, J.-Y. (1995) J. Amer. Soc. Brew. Chem., 53, 157.<br />
SCOTT, R. W. (1972) J. Inst. <strong>Brewing</strong>, 78, 179, 411.<br />
SCOTT, P. M. (1996) J. A.O.A.C. International, 79, 875.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
SISSONS, M. J. (1996) J. Amer. Soc. Brew. Chem., 54, 19.<br />
SJOÈ HOLM, K., MACRI, L. J. <strong>and</strong> MACGREGOR, A. W. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels,<br />
p. 277.<br />
SOPANEN, T., TAKKINEN, P., MIKOLA, J. <strong>and</strong> ENARI, T.-M. (1980) J. Inst. <strong>Brewing</strong>, 86, 211.<br />
STARS, A. C., SOUTH, J. B. <strong>and</strong> SMITH, N. A. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo, p. 103.<br />
STENHOLM, K. <strong>and</strong> HOME, S. (1999) J. Inst. <strong>Brewing</strong>, 105, 205.<br />
STENHOLM, K., HOME, S., PIETILaÈ, K., MACRI, L. J. <strong>and</strong> MACGREGOR, A. W. (1996) Proc. 24th Conv. Inst.<br />
<strong>Brewing</strong>, (Asia Pacific Section), Singapore, p. 142.<br />
STONE, B. A. (1996) in Cereal Grain Quality (Henry, R. J. <strong>and</strong> Kettlewell, P. S. eds). London, Chapman &<br />
Hall, pp. 251, 288.<br />
SUN, Z. <strong>and</strong> HENSON, C. A. (1990) Plant Physiol., 80, 310.<br />
TAYLOR, D. G. (1981) Brew. Distill. Internat., 11 (4), 35, 42.<br />
TAYLOR, L. (1974) Eur. Brew. Conv. Monograph-I. E. B. C.-Wortsymposium, Zeist, p. 208.<br />
TENG, J., STUBITS, M. <strong>and</strong> LIN, E. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 47.<br />
TESTER, R. F. (1997) in Starch Structure <strong>and</strong> Functionality (Frazier, P. J., Richmond, P. <strong>and</strong> Donald, A. M.<br />
eds). Cambridge. Royal Society of Chemistry, p. 163.<br />
VAN DEN BERG, R. <strong>and</strong> VAN EERDE, P. (1982) Proc. 17th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong><br />
Sect.), Perth, p. 70.<br />
VOSE, J. R. (1979) M. B. A. A. Tech. Quart., 16, 186.<br />
WACKERBAUER, K., ZUFALL, C. <strong>and</strong> HOÈ LSCHER, K. (1993) Brauwelt Internat., 11, 107.<br />
WAINWRIGHT, T. (1980) Eur. Brew. Conv. Monograph-I. E. B. C.-Symposium on the relationship between<br />
malt <strong>and</strong> beer, Helsinki, p. 118.<br />
WALKER, M. D., BOURNE, D. T. <strong>and</strong> WENN, R. V. (1997) Proc. 26th Congr. Eur. Brew. Conv., Maastricht,<br />
p. 191.<br />
WEBSTER, R. (1978) Brewer's Guard., 107 (7), 51, 56.<br />
WHITEAR, A. L., MAULE, D. R. <strong>and</strong> SHARPE, F. R. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 81.<br />
WHITTLE, N., ELDRIDGE, H., BARTLEY, J. <strong>and</strong> ORGAN, G. (1999) J. Inst. <strong>Brewing</strong>, 105, 89.<br />
WIEG, A. J. (1987) in <strong>Brewing</strong> Science 3 (Pollock, J. R. A., ed.). London. Academic Press, p. 533.<br />
WINDISCH, W., KOLBACH, P. <strong>and</strong> SCHILD, E. (1932) Wochensch. f. Brau., XLIX, (37, 38), 289, 298.<br />
WOOF, J. B. <strong>and</strong> PIERCE, J. S. (1966) J. Inst. <strong>Brewing</strong>, 72, 40.<br />
WRIGHT, H. E. (1892) A H<strong>and</strong>ybook for Brewers. London. Crosby Lockwood, 516 pp.<br />
YAMADA, K. <strong>and</strong> YOSHIDA, T. (1976) Rept. Res. Lab. Kirin Brewery Co., no. 19, 25.<br />
ZHANG, N. <strong>and</strong> JONES, B. L. (1995a) J. Cereal Sci., 21, 145.<br />
ZHANG N. <strong>and</strong> JONES, B. L. (1995b) J. Cereal Sci., 22, 147.<br />
ZIEGLER, L. <strong>and</strong> PIENDL, A. (1976) M. B. A. A. Tech. Quart., 13, 177.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
5<br />
The preparation of grists<br />
5.1 Intake, h<strong>and</strong>ling <strong>and</strong> storage of raw materials<br />
Malts<strong>and</strong>mashtunadjunctsmaybedeliveredbyroad,railor,morerarely,water.Forlarge<br />
breweries deliveries are in bulk, but for smaller ones deliveries may be in sacks varying in<br />
weight from about 25kg. to 1t. The largest sacks are h<strong>and</strong>led with fork-lift trucks. Lorries<br />
generally carry loads of 20±25t, while American railway wagons carry about 68±82tof<br />
malt. Some small breweries receive their malt ready ground. Before aload is accepted it<br />
shouldbecheckedtoensurethatitisofthecorrectquality.Visualinspection<strong>and</strong>rapidtests<br />
onmaltshouldindicatethatthecolour,moisturecontent,nitrogen(protein)content,flavour,<br />
aroma, friability, homogeneity <strong>and</strong> range of corn sizes are correct <strong>and</strong> that the load is free<br />
from insects <strong>and</strong> the corns are not unduly damaged (Briggs, 1998; Kunze, 1996; Narziss,<br />
1992; Nicol <strong>and</strong> Andrews, 1996; Rehberger <strong>and</strong> Luther, 1994; Sugden et al., 1999).<br />
Different types of malt <strong>and</strong> adjuncts must not be allowed to mix, either during h<strong>and</strong>ling<br />
or subsequent storage. Each load is weighed, for example, by weighing each lorry on a<br />
weighbridge before <strong>and</strong> after unloading, <strong>and</strong> during movements within the brewery. While<br />
mechanical h<strong>and</strong>ling can be used for malts <strong>and</strong> many adjuncts, flours must be moved using<br />
pneumatic equipment. Many different arrangements of equipment are used in dry-goods<br />
h<strong>and</strong>ling (Fig. 5.1). In the UK lorries usually discharge by tipping their loads into a<br />
reception pit which is sheltered from the weather <strong>and</strong> is aspirated to remove dust.<br />
H<strong>and</strong>ling is with equipment that damages the material being moved as little as<br />
possible <strong>and</strong> so causes minimal losses <strong>and</strong> generates least dust. Usually the machines<br />
used are belt <strong>and</strong> bucket elevators, <strong>and</strong> screw, drag, chain-<strong>and</strong>-flight or en masse<br />
conveyors (Briggs, 1998). All these should be aspirated to remove dust, which is usually<br />
collected at a central point. Initially, malt is often roughly screened (sieved) to remove<br />
coarse <strong>and</strong> fine impurities, <strong>and</strong> is passed over magnetic separators (of fixed or revolving<br />
magnet types) to remove fragments of metal. Sometimes the malt is separated from<br />
`heavy contaminants' by passing it through a transverse airflow, which deflects the<br />
comparatively light malt while allowing denser objects to continue falling downward, to<br />
be collected separately. The removal of metal items (`tramp iron') <strong>and</strong> subsequent destoning<br />
are necessary to reduce wear on conveying equipment <strong>and</strong> the brewery mills <strong>and</strong><br />
to reduce the risks of sparks which can lead to fires or explosions.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Blower<br />
Malt<br />
silo<br />
Adjunct<br />
silo<br />
Rotary seal<br />
Malt<br />
bins<br />
Screen<br />
Rejected<br />
material<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Elevator<br />
Magnet<br />
Dust<br />
Fan<br />
Screened<br />
malt<br />
Dust<br />
Adjunct blowline<br />
Waste<br />
Filter<br />
unit<br />
Rotary<br />
seal<br />
Tip<br />
weigher<br />
De-stoner<br />
Buffer bin<br />
Steam<br />
Fig. 5.1 An example of a dry goods h<strong>and</strong>ling system.<br />
Waste<br />
Adjunct<br />
grist case<br />
Fan<br />
Cyclone<br />
Conditioning<br />
screw<br />
Mill<br />
Grist conveyor<br />
To cereal<br />
cooker or<br />
mashing vessel<br />
Waste<br />
To mash<br />
vessel<br />
Air<br />
Filter<br />
unit<br />
Malt<br />
grist<br />
case
In British breweries malt, as now delivered, often meets such tight specifications that<br />
preliminary screening is not needed. Grist materials, <strong>and</strong> especially malts, are<br />
hygroscopic <strong>and</strong> efforts are made during h<strong>and</strong>ling <strong>and</strong> storage to prevent them picking<br />
up moisture by contact with water or from the air <strong>and</strong> so becoming `slack'. Pneumatic<br />
conveyors may function on positive or negative (`suction') pressures. In positive-pressure<br />
systems a blower forces air along fixed or flexible pipework <strong>and</strong> malt or other material is<br />
introduced into the airstream from hoppers, via air-tight, rotating valves. The material is<br />
swept along <strong>and</strong> is recovered from the airstream in expansion chambers <strong>and</strong> cyclones.<br />
The material drops out from the slower airflow into a receiving hopper. Dust is separated<br />
from the air, using cyclones <strong>and</strong> textile filters, before escaping from the building. Modern<br />
filters are self-cleaning <strong>and</strong> occasional backflows of air dislodge the dust caked onto the<br />
filter sleeves, allowing it to fall into a reception hopper, from which it is removed through<br />
a valve. Negative-pressure systems work under suction so that material is sucked into the<br />
system. This is convenient for unloading rail wagons <strong>and</strong> barges <strong>and</strong> is also advantageous<br />
in that dust does not leak <strong>and</strong> escape. This type of conveyor can form part of the main<br />
dust collection system. The `used' air passes through cyclones <strong>and</strong> dust filters before it<br />
reaches the pump that provides the suction. Some grain-cleaning equipment re-circulates<br />
air, preventing the dust escaping <strong>and</strong> collecting it in the machine.<br />
H<strong>and</strong>ling dry goods always generates dust. This is heavily infested with microbes <strong>and</strong><br />
their spores <strong>and</strong>, mixed with air in a range of proportions, is highly explosive. It is<br />
estimated that 0.4±1.4% of malt delivered to a brewery is damaged <strong>and</strong> converted into dust.<br />
Intake hoppers, conveyors, elevators, stores, screens, de-stoners <strong>and</strong> dry mills should all be<br />
ventilated <strong>and</strong> the aspirating air directed to the dust-collection system. The dust is usually<br />
mixed with the spent grains <strong>and</strong> sold. Less usually it may be discarded or destroyed. Dust<br />
from malt <strong>and</strong> adjuncts is rich in starch <strong>and</strong> may be added into mashes. Dust, either in or<br />
around equipment such as mills <strong>and</strong> conveyors, should not be allowed to accumulate as it<br />
becomes damp <strong>and</strong> microbes <strong>and</strong> insects multiply on it. Dust must be kept away from the<br />
brewing plant as the microbes in it can initiate beer spoilage. In addition it constitutes a<br />
health risk, giving rise to skin allergies <strong>and</strong> lung infections. Dust explosions can be highly<br />
destructive <strong>and</strong> life-threatening. Rigid rules of behaviour <strong>and</strong> the use of well-designed<br />
equipment must be used to minimize the risks. Dust containment <strong>and</strong> removal are essential.<br />
Explosions may be initiated by sparks or flames <strong>and</strong> so stones <strong>and</strong> tramp iron are removed<br />
from malt <strong>and</strong> adjuncts, smoking <strong>and</strong> the use of flames (e.g. welding torches) is forbidden<br />
<strong>and</strong> all equipment is earthed or `grounded' to discharge static electricity safely <strong>and</strong> avoid<br />
sparking. Sensors are mounted on bearings to detect increases in temperature, which<br />
indicate poor lubrication, wear <strong>and</strong> the risk of initiating a dust explosion.<br />
Overloaded equipment is automatically closed down. Silos have dust-sealed lighting <strong>and</strong><br />
silos <strong>and</strong> ducts are fitted with physically weak explosion relief hatches that will burst on a<br />
sudden pressure rise <strong>and</strong> will allow the explosion to vent in a comparatively safe way. This<br />
reduces the force being transmitted through the ducting <strong>and</strong> so reduces the chance of<br />
secondary explosions <strong>and</strong> other damage remote from the initial explosion site. Some pieces<br />
of equipment, such as dry mills, are fitted with devices which, when triggered, release an<br />
inert gas <strong>and</strong> quench explosion flames. The three most common sites of explosions in<br />
breweries, in decreasing order, are bucket elevators > silos > mills. As with all equipment,<br />
the dust-collection system should be inspected regularly, <strong>and</strong> be cleaned <strong>and</strong> maintained.<br />
Malts <strong>and</strong> adjuncts are stored so that they do not become damp <strong>and</strong> `slack' <strong>and</strong> they<br />
are protected from birds, rats, mice <strong>and</strong> insects. Storage is costly, <strong>and</strong> so it is desirable to<br />
have as small stocks as are considered prudent. In <strong>practice</strong>, breweries may carry stocks<br />
sufficient to allow from three days' to three weeks' production. In smaller or older<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
eweries themalts<strong>and</strong>adjunctsmaybestored insacks,traditionallyinaroomabovethe<br />
mill room. Usually malt is stored in bulk in bins or silos. These are usually of steel or<br />
concrete,withsmoothinternalsurfaceswithconicalbaseshavingavalveat thebottomto<br />
allow metered unloading, under gravity, into a conveyor. Different materials have<br />
different angles of repose <strong>and</strong> they need different valley angles in the tapered or conical<br />
bases to allow them to flow freely (see Appendix).<br />
There is atendency for flowing, granular material to segregate by size <strong>and</strong> density. To<br />
avoid this stores should be unloaded using `bulk flow' systems that avoid funnel flow.<br />
This can be achieved using suitable vessel designs (Farnish, 2002). Flour is often slow to<br />
flow, <strong>and</strong>itmay benecessarytohave avibrating cone oramechanical device (suchasan<br />
arch-breaker) in the cone to encourage movement. Silos are loaded by conveyor from<br />
above <strong>and</strong> are equipped with hatches <strong>and</strong> manways to allow inspections <strong>and</strong> cleaning.<br />
Level-detecting devices determine the approximate levels of the contents <strong>and</strong> prevent<br />
over-filling. They should contain thermometers to detect any temperature rises that give<br />
warning of the onset of spoilage. It is good <strong>practice</strong> to use old stocks first <strong>and</strong> never to<br />
mix old <strong>and</strong> new stocks. Each storage container should be completely emptied in turn,<br />
<strong>and</strong> be cleaned <strong>and</strong>, if necessary, fumigated before being refilled. If grist materials must<br />
be treated to control insects this must be done using only fully trained staff, using an<br />
approved insecticide or fumigant at apermitted dose rate.<br />
Batchesofmaltsoradjuncts<strong>and</strong>gristsareweighedastheyaremovedaroundinthestore,<br />
or to or from the mills. The records of the weighings are used to check on losses, that the<br />
correctmixturesofmaterialsentereachmash,<strong>and</strong>thattheyieldofextractfromeachmashis<br />
acceptable. Electronic <strong>and</strong> mechanical weighers <strong>and</strong> load cells are in use <strong>and</strong> some are<br />
capable of high degrees of accuracy. In many weighers the malt, or other material, flows<br />
from above into a weighing container. When the chosen weight is approached the flow-rate<br />
is reduced <strong>and</strong> as the weight is reached the flow is stopped. The weighing container then<br />
empties from the base into a receiving buffer hopper. When it is empty the base closes <strong>and</strong><br />
refilling is resumed. Weighing is recorded as the number of fillings that have taken place.<br />
Sometimes hoppers <strong>and</strong> other containers, such as grist cases, are mounted on load cells,<br />
which indicate the weight of the contents. Load cells can be mounted under conveyors to<br />
measure the weight of material continuously as it is being conveyed. While older weighers<br />
were perhaps accurate to 2% new machines, using load cells, can be good to 0.1%.<br />
Before the grist materials reach the mill they may be screened again, removing all<br />
items larger or smaller than malt corns. They will pass over a magnetic separator <strong>and</strong> they<br />
should pass through a de-stoner <strong>and</strong>/or a `heavy object separator' <strong>and</strong> be aspirated. In destoners<br />
a thin stream of malt flows slowly onto the upper end of an inclined screen. Air<br />
passes up through the screen at such a rate that the malt `floats' <strong>and</strong> flows downwards,<br />
being collected at the base. Heavier objects, such as stones, are not lifted <strong>and</strong> rest on the<br />
screen. This has a jerking motion that moves the stones up the screen <strong>and</strong> over the upper<br />
edge into a collector (Briggs, 1998).<br />
When mixtures of different grist materials are to be used they must be moved to the mills<br />
<strong>and</strong> the grist case, bypassing the mill in some cases (as when wheat flour is used), in the<br />
correct proportions <strong>and</strong> at the correct rates so that the correct grist, well mixed, is supplied<br />
when mashing in occurs. `Dry goods h<strong>and</strong>ling' is controlled from a central point, where the<br />
states of the stores, whether equipment is running <strong>and</strong> the states of processes are monitored<br />
<strong>and</strong> controlled often with a mimic display or, increasingly, through a computer. Each system<br />
has built-in safety <strong>and</strong> warning devices (Bhaduri, 1996). For example, when the flow of malt<br />
from a silo to the mill is switched off the valves, conveyors, elevators, etc., are switched off<br />
in sequence, beginning at the silo, so each unit is emptied of malt before it stops running.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Start up is in the reverse order. For safety <strong>and</strong> effective operation all the staff must be well<br />
trained.Checking<strong>and</strong>maintenanceoftheequipmentmustbecarriedoutregularly.Safetyis<br />
also dependent on the layout of the plant <strong>and</strong> this <strong>and</strong> the way it is operated must meet high<br />
design st<strong>and</strong>ards <strong>and</strong> local regulations.<br />
5.2 The principles of milling<br />
The objective ofmilling istobreak up malt<strong>and</strong>adjuncts tosuchanextent thatthegreatest<br />
yield of extract is produced in the shortest time in the mashing equipment in use. With<br />
most wort separation systems it is desirable to keep malt husk as intact as possible, to help<br />
maintain an open filter bed that favours wort separation. No unbroken grains should<br />
survive milling. Only comparatively coarse grists can be used in mash tuns. If the grist is<br />
too fine (has too high aproportion of small particles) the wort will not separate from the<br />
spent grains. Finer grists can be used in lauter tuns <strong>and</strong> still finer grists in mash filters. A<br />
`best practical' grist, with an optimal mixture of coarse, fine <strong>and</strong> very fine particles, is<br />
determined for each mashing system. Optimal milling is not simple to achieve. Mill<br />
settings need to be altered for adjuncts or malts differing in their degrees of modification<br />
(badly modified malts should be milled to afiner grist, well modified malts readily<br />
shatter), <strong>and</strong> quite small differences in moisture contents alter the effectiveness of milling<br />
using one mill setting (Crescenzi, 1987). Sometimes different mills are used for malt <strong>and</strong><br />
for adjuncts, or for pale malts <strong>and</strong> coloured malts. In other cases adjuncts, such as torrified<br />
barley or wheat, flaked cereals, <strong>and</strong> roasted barley, may be milled mixed with the malt(s).<br />
The term `grinding', for milling, comes from the days when malt was ground between<br />
millstones, <strong>and</strong> the ground material is still termed `grist'. This should be distinguished<br />
from grain particles, present in grists, which are termed `grits'. Milling systems may use<br />
rollermilling,impactmillingwithdisc-orhammer-mills,<strong>and</strong>wetmilling.Properly,none<br />
of these modern mills `grinds' the malt. Roller milling (dry or conditioned), giving<br />
comparatively dry grists, is the most usual. These grists are evaluated by inspection, by<br />
sieve analyses <strong>and</strong> sometimes in laboratory mashes. A grist should be uniform in<br />
appearance, be free from taints <strong>and</strong> insects or other contaminants <strong>and</strong> should not contain<br />
any whole grains or large grain pieces that indicate that part of the grist has not been<br />
effectively milled.<br />
Sieve analyses should be used regularly to check that amill's performance is not<br />
drifting. The most commonly used sets of sieves are the Pfungstadt plansifter (EBC), the<br />
MEBAK sieves <strong>and</strong> the sieves of the ASBC (Table 5.1). Asample of grist is loaded onto<br />
the top of aset of horizontal sieves, which is shaken mechanically for afixed period of<br />
time. The percentage, by weight, of the grist retained on each, successively finer sieve is<br />
then determined. By convention the sieve fractions are given names (Table 5.1), but the<br />
fractions are not `pure' so, for example, the `husk' fraction contains tissues from the rest<br />
of the malt grains <strong>and</strong> husk tissue occurs in the other fractions. Furthermore, the grist<br />
fractions from amalt milled in different ways are likely to have different compositions.<br />
The results in Table 5.2 show that starch is present in the `husk' fractions proving<br />
contamination with endosperm material while the distribution of fibre in all the fractions<br />
suggests that at least traces of husk occur in all of them. The variable extract yields <strong>and</strong><br />
qualities obtained when different fractions of agrist are mashed has often been noticed<br />
(Hind, 1950; Narziss, 1992; Stubits et al., 1986; Table 5.3). It has been suggested that<br />
different beers could be produced by combining the fractions in different proportions in<br />
the mash (e.g. Isoe et al., 1991), or that beers should be made with husk-depleted grists or<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 5.1 Details of the sieves used to characterize grists. The names given to the fractions do<br />
NOT indicate that these are morphologically pure grain fractions. The composition of the fractions<br />
changes with the type of milling employed. The American Society of <strong>Brewing</strong> Chemists (ASBC)<br />
<strong>and</strong> the European Brewery Convention (EBC) (Pfungstadt) sieves are significantly different <strong>and</strong> the<br />
MEBAK sieves are different again (see footnote). All dimensions are in millimetres<br />
ASBC EBC<br />
Sieve Mesh Fraction Sieve Wire Mesh Fraction<br />
number width name (Pfungstadt) thickness width name<br />
(mm) number (mm) (mm)<br />
10 2.000 Husk 1 (16) 0.31 1.270 Husk<br />
14 1.410 Husk 2 (20) 0.26 1.010 Coarse grits<br />
18 1.000 Husk 3 (36) 0.15 0.547 Fine grits I<br />
30 0.590 Coarse grits 4 (85) 0.07 0.253 Fine grits II<br />
60 0.250 Fine grits 5 (140) 0.04 0.152 Flour<br />
100 0.149 Flour Tray (through) ± ± Fine flour<br />
Pan (through) ± Fine flour<br />
The MEBAK screens I, II, III, IV <strong>and</strong> V have mesh widths (mm) of 1.25, 1.00, 0.5, 0.25 <strong>and</strong> 0.125.<br />
Table 5.2 The analyses of ASBC sieve fractions of Larker malt milled with a 5-roller BuÈhler-<br />
Miag malt mill (data of Stubits et al., 1986). Grist samples (100 g) were mashed with a liquor/grist<br />
ratio of 3/1. The mash programme was 45 ëC/67 min., then a temperature rise of 1 ëC/min. to 69 ëC,<br />
a hold for 35 min., then a rise to 72 ëC<br />
Sieve Starch Protein Fibre Wort Viscosity<br />
number (%) (%) (%) (ëP) (relative)<br />
10 50.1 12.0 7.1 16.6 2.07<br />
14 56.1 13.8 4.1 19.1 2.60<br />
18 63.4 11.6 3.5 20.9 2.62<br />
30 54.9 13.4 3.7 20.4 2.50<br />
60 47.1 15.4 5.2 19.8 2.44<br />
100 49.4 17.7 4.3 20.1 2.41<br />
Pan 69.9 11.8 1.9 21.7 2.54<br />
Table 5.3 Analyses of a malt grist <strong>and</strong> the Pfungstadt sieve fractions 1±6 of that grist (data of<br />
Narziss <strong>and</strong> Krauss, via Narziss, 1992)<br />
Analyses Entire Husks Coarse Fine Fine Flour Fine<br />
grist (1) grits (2) grits I (3) grits II (4) (5) flour (6)<br />
Proportions (%) 100 27.6 15.3 22.9 13.2 6.6 14.4<br />
Extract (%)* 80.2 64.4 79.5 87.9 84.2 83.3 96.8<br />
Saccharification<br />
time (min.) 9 8 9 8 9 10 12<br />
Fermentation<br />
limit (%) 80.9 77.3 78.0 82.0 82.5 80.9 83.2<br />
Protein (%)* 11.1 12.4 11.9 10.6 11.4 13.4 7.6<br />
TSN (mg/100 g)* 711 584 681 705 847 854 526<br />
Viscosity (mPa.s,<br />
8.6% wort) 1.515 1.534 1.463 1.481 1.443 1.467 1.407<br />
D P (ëW.-K) 302 225 323 361 347 327 250<br />
-Amylase (ASBC) 40 32 44 51 48 47 36<br />
Tannoids (mg PVP/<br />
100 g)* 22 12 32 20 21 24 12<br />
Colour (ëEBC) 3.3 4.7 2.8 2.5 2.8 2.8 1.3<br />
TSN, total soluble nitrogen. D P, diastatic power. PVP, polyvinyl pyrrolidone. * Dry weight basis.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 5.4 EBC/Pfungstadt sieve analyses (% in each fraction) of all-malt grists suitable for<br />
different mashing systems (Kunze, 1996; Narziss, 1992; Sugden et al., 1999)<br />
Sieve Mash tun Lauter tun Mash filter (conventional) Thin-bed filter<br />
a b a b c a b<br />
1 27 18 18±25 11 8±12 10±12 1 1<br />
2 9 8 < 10 4 3±5 4±6 4 2<br />
3 24 35 35 16 15±25 < 15 9 15<br />
4 18 21 21 43 35±45 > 43 26 29<br />
5 14 7 7 10 8±11 > 12 19 24<br />
Tray 8 11 < 15 16 12±18 < 16 41 29<br />
the fractions should be mashed separately in different ways, to obtain abeer with a<br />
`refined' flavour (Krottenthaler et al., 1999). While this is technically possible this<br />
approach has been little used, probably because of the extra cost of the fractionation<br />
procedures <strong>and</strong> the problem of using or disposing of the `husk' fraction.<br />
Examples of all-malt grists regarded as being suitable for use with different mashing<br />
systems are given in Table 5.4. In moving from the mash tun grist across to the grist for the<br />
thin-bed filter the grists become finer. Itis seen thatthere are small differences between the<br />
proposed grists. This is expected as `best' grists are determined by brewers using trial <strong>and</strong><br />
error, employing different mills <strong>and</strong> equipment. Grists prepared by wet milling are not<br />
suitable for particle size analyses. In these cases grist quality is judged indirectly by noting<br />
the appearance <strong>and</strong> performance of mashes, particularly extract yield, run-off rate, wort<br />
clarity, the appearance of the spent grains <strong>and</strong> so on. The same criteria are also applied to<br />
mashes made with dry milled grists, which are also subjected to sieve analysis.<br />
Roll (or roller) mills are commonly used in breweries. The rollers work in pairs. The<br />
malt grains are delivered to the first pair of rolls by afeed roll that determines the feed<br />
rate<strong>and</strong>isintendedtodelivereachcorn`end-on' totheworkingrolls.Each cornisdrawn<br />
between the rolls <strong>and</strong> is crushed, sheared (if the rolls are rotating at different speeds) <strong>and</strong><br />
cut if the rolls are fluted (grooved). The theoretical capacity of aroll mill, Q(m 3 /h), is<br />
given as Qˆ60.s.NL10 9 ,where sˆrotational speed (rpm); Nˆthe gap between the<br />
rollers or `nip'(mm); Lˆthe length of the working surfaces of the rolls (mm). In fact the<br />
practical working capacity is 10±30% of the theoretical capacity (Sugden et al., 1999).<br />
Working rates of different mills (in kg/h/mm roll length) are given as two-roll mills, 1.5±<br />
2.5; four-roll mills, 2±6 <strong>and</strong> six-roll mills, 1.5±10.<br />
Theactionofcrushingrolls,withcentresA1<strong>and</strong>A2,onaparticle,centre B,isillustrated<br />
in Fig. 5.2. Acomponent, e, of the force ttends to draw the particle between the rolls. The<br />
force t depends on the force r <strong>and</strong> the coefficient of friction between the surfaces of the<br />
particle <strong>and</strong> the roll, , so t ˆ r. The force components e <strong>and</strong> m are opposed <strong>and</strong> so unless<br />
e > m the particle will not be pulled between the rolls <strong>and</strong> be crushed. Thus r cos > r<br />
sin <strong>and</strong> so > tan which must, therefore, be less than the coefficient of friction. Often<br />
a typical value for is 16 ë. The angle OEF, the angle of nip, equals 2 There is a definite<br />
relationship, cos ˆ radius of the roll + half the gap between the rolls)/(radius of the roll +<br />
radius of the particle). If is 16 ë, then cos ˆ 0.961. If the angle of nip is too large<br />
(because the rollers are too small) <strong>and</strong>/or the rolls are rotating too quickly, particles `ride<br />
the rolls', so they are not drawn between them <strong>and</strong> they are not crushed.<br />
The peripheral roll speeds in brewery mills are often 2.4±4 metres/sec (8±13 ft./sec.). As a<br />
particle moves between the rollers so it deforms <strong>and</strong>, if it is brittle, its structure fails <strong>and</strong> it<br />
breaks up (Sugden et al., 1999). Rolls may move at different speeds, for example the faster<br />
may rotate at 1.25 the speed of the slower. Consequently a particle passing between them<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
m<br />
A 1<br />
r<br />
α<br />
n<br />
will be torn by shear as well as being crushed. As well as that the grooves, or `fluting' milled<br />
in spirals on the surfaces of the rollers can be arranged not only to increase the coefficient of<br />
friction between the particles <strong>and</strong> the rolls, but also to cut the particles (Kunze, 1996).<br />
Roll mills differ in their complexity. They have to cope with materials, particularly malt<br />
corns, that differ in their ranges of widths <strong>and</strong> so the nips, the gaps between the rolls, must be<br />
adjustable. In addition corns <strong>and</strong> other materials break up to give particles of a range of sizes.<br />
Some of these, the fine grits <strong>and</strong> flour, need no more milling while larger materials can, with<br />
advantage, be broken up more. In more complex mills this is achieved by sieving the material<br />
coming from a pair of rolls on cylindrical or flat-bed screens. The fine fractions leave the mill<br />
directly while coarser fractions are directed between other pairs of rolls <strong>and</strong> are broken up<br />
further. In the cylindrical screens the grist may be disrupted by spinning `beaters'. In the flatbed,<br />
oscillating screens rubber balls move about <strong>and</strong> oscillate vibrating the sieving surfaces<br />
<strong>and</strong> keeping them clear of blockages. Dry mills deliver their grist to a receiving hopper or<br />
`grist case', so several hours working time may be available to prepare a batch of grist. In<br />
contrast, with `wet' mills the grist is mashed in as it is produced <strong>and</strong> the mashing-in period<br />
must not be too long. Consequently wet mills operate at faster rates than dry mills.<br />
5.3 Laboratory mills<br />
O<br />
f<br />
e α t<br />
D<br />
C<br />
α α<br />
E<br />
B<br />
Fig. 5.2 The action of two crushing rolls, rotating at equal speeds, on a particle such as a malt<br />
corn (see the text).<br />
Many kinds of mills are used in laboratories, but for st<strong>and</strong>ard analyses particular mills,<br />
operated in closely defined ways, are employed. This is because the analyses are used as<br />
bases of commercial transactions as well as st<strong>and</strong>ards by which the value of malts to a<br />
brewery can be judged <strong>and</strong> so different laboratories must obtain analytical results that are in<br />
close agreement. During the last century the `st<strong>and</strong>ard' mills have changed. Initially h<strong>and</strong>cranked<br />
Boby or Seck mills, each with one pair of small rolls, were used <strong>and</strong>, in time, these<br />
were powered with electric motors. These mills were not able to give sufficiently<br />
reproducible results <strong>and</strong> so, at least for fine grinds, some used cone mills. The EBC<br />
introduced a Casella mill in which the malt was shattered by the blades on a spinning rotor<br />
<strong>and</strong> the grist was collected after passing through a sieve, different sieves being used for the<br />
coarse <strong>and</strong> fine grinds. In a comparatively short time this mill was replaced by another, a<br />
BuÈhler-Miag disc mill, which is in use at present by both the EBC <strong>and</strong> the IGB. In this<br />
sophisticated machine the sample to be ground is fed into the central area between two<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
F<br />
A 2
discs, one static, the other rotating fast at afixed speed. The grist emerges from the gap at<br />
theperiphery<strong>and</strong>iscollected. The fineness ofgrindisdeterminedbythedistancebetween<br />
thediscs,whichfortheIoBmethodsare0.7mmforthecoarsegrind<strong>and</strong>0.2mmforthefine<br />
grind. It is apparent that this mill differs in its method of working from almost all<br />
commercial malt mills used in brewing. This `disadvantage' is more apparent that real,<br />
since the results it gives are highly reproducible <strong>and</strong>, as with every analytical system, each<br />
brewer has to establish the relationships between the laboratory analyses <strong>and</strong> the results<br />
obtained with his, possibly unique, brewing system.<br />
5.4 Dry roller milling<br />
Alarge number of types ofroller millshave been, or are,in use (Kunze, 1996; Narziss, 1992;<br />
Sugdenet al.,1999). `Dry' millshave some characteristics in common. Thefeed roll delivers<br />
themalttothefirstpairofcrushingrolls,acrosstheirfullwidth,atacontrolledrate.Therolls<br />
aredesignedtodeliverthecorns`endon'tofavourtheirbeingcrushedalongtheirlengthwith<br />
theminimumdegreeofhuskbreakage.Therollsareoftenabout250mm(9.84in.)indiameter<br />
(inwetmillstheyareoftenlarger).Asnotedtherollsareusuallyfluted<strong>and</strong>mayrunatdifferent<br />
speeds. Bothmay be drivenbutsometimes onlyone ispowered, the`follower' being dragged<br />
bythefriction between thegrist<strong>and</strong>themoving,poweredroll. Millrollsmayoperate at 250±<br />
500rpm. The roll length increases with machine capacity, to amaximum of about 1500mm<br />
(59.06in.). The rolls are spring loaded, <strong>and</strong> so the gaps between them must be checked while<br />
theyareworkingunderload.Thisisachievedbypassingsoftmetalwirebetweentheworking<br />
rolls,thenmeasuringthethicknessofthesquashedmetalwithamicrometerscrewgauge.Lead<br />
wire was originally used, but now aluminium, copper or lead-free solder wires are employed,<br />
toavoidanychanceofleadcontamination.Therollgaps<strong>and</strong>thealignmentoftherollsmustbe<br />
checked regularly, to ensure that they are parallel. In addition the rolls must be checked for<br />
wear. The grist coming from each section of the roll pairs must be inspected <strong>and</strong> checked by<br />
sieveanalysis.Thistestwilldetectwhenthecentralportionsoftherollsarebadlywornbutthe<br />
endregionsarenot.Wherescreensareusedthesetoomustbecheckedforintegrity.Afteruse<br />
the mills should be cleaned. At no time should insect infestations be present.<br />
The simplest mills are single-pass, two-roll dry mills. These are relatively slow<br />
working <strong>and</strong> are inflexible, being suitable only for well-modified malts, special malts or<br />
rice. They are used only in small units, such as pub breweries. Gaps used are 0.6±1.0mm<br />
(0.024±0.039in.) <strong>and</strong> working capacities are 1.5±2.5kg/h/mm roll length (a working rate<br />
of 1kg/h/mm roll length is almost exactly 60lb/h/in. roll length). It seems that three-roll<br />
mills (with or without screens) are no longer in use. The arrangement of the three rolls<br />
was like the grouping found in five-roll mills (see below; Sugden et al., 1999).<br />
Four-roll(two-high)millsarecommoninsmaller,traditionalalebreweries(Figs5.3<strong>and</strong><br />
5.4). They are robust, relatively inexpensive <strong>and</strong> well suited for milling well-modified<br />
malts with a small range of corn sizes. Working rates are 2±6 kg/h/mm, the gaps between<br />
the upper rolls are 1.3±1.9 mm (0.051±0.075 in.) <strong>and</strong> between the lower rolls 0.3±1.0 mm<br />
(0.012±0.039 in.). In the more simple types the cracked grist from the first rolls piles up on<br />
a `dam', `shelf' or `explosion preventer', where it can smother any sparks <strong>and</strong> so prevent an<br />
explosion (Fig. 5.3). It then spills over <strong>and</strong> falls into a beater chamber, where the crushed<br />
malt is broken up <strong>and</strong> flour <strong>and</strong> grits are partly separated from the husk before passing<br />
through the second pair of rolls, over a second explosion preventer <strong>and</strong> out to the grist case.<br />
More sophisticated types of four-roll mills use screens to separate the grist into fractions of<br />
which only the coarse grits are crushed further by passage between the second pair of rolls<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fixed roll (fluted)<br />
Inspection door<br />
Sample valve<br />
H<strong>and</strong>le operating<br />
valve for sample<br />
for top rails<br />
Inspection door<br />
Fixed roll (plain)<br />
Sample shoot<br />
Feed inspection<br />
door<br />
Sample<br />
tube<br />
Feed valve<br />
Explosion<br />
preventer<br />
Feed roll<br />
Fixed roll<br />
(fluted)<br />
Explosion<br />
preventer<br />
Beaters<br />
Loose roll<br />
(plain)<br />
Fig. 5.3 Asimple four-roll mill (after Hind, 1950). Note that in the second pair of rolls one is<br />
driven (`fixed') while the other is not powered (`loose').<br />
(Fig.5.4).Thisavoidsmoredamagetothehuskmaterial<strong>and</strong>sofavoursagoodwortrun-off<br />
rate. The screens may be semi-cylindrical <strong>and</strong> equipped with revolving beaters, as in Fig.<br />
5.4, or they may be oscillating, flat-bed screens. The larger the screen areas the better the<br />
separations achieved. The arrangements of the screens can be varied so that, for example,<br />
flour<strong>and</strong>gritsarere-milledbuthuskisnotor,alternatively,grits<strong>and</strong>huskarere-milledbut<br />
flour <strong>and</strong> fine grits are not (Sugden et al., 1999).<br />
So-called eight-roll mills were really two four-roll mills working in parallel. The malt<br />
was pre-screened to divide it into bold (plump) <strong>and</strong> thin corns <strong>and</strong> these were milled<br />
separately, each stream passing through afour-roll mill adjusted to deal with the particular<br />
cornsizes.Five-rollmillsarenowuncommon(Fig.5.5),buttheyarelessexpensivethatthe<br />
six-roll mills. The malt is crushed between the first pair of rolls (say 1<strong>and</strong> 2) <strong>and</strong> the grist<br />
fromthisoperationisscreened.Thehuskfractioniscrushedfurther,(betweenrolls2<strong>and</strong>3)<br />
<strong>and</strong> the products are separated on asecond set of screens which direct the flour <strong>and</strong> husks<br />
(withgritsattached)toleavethemill,<strong>and</strong>thegritstopassthroughthethirdpairofrolls(say<br />
4<strong>and</strong>5).Atthesametimetheflourfromthefirstpairofrollsisguidedoutofthemill<strong>and</strong>the<br />
grits join those from the second pair of rolls <strong>and</strong> are broken up further by the third pair.<br />
Six-roll mills are widely used in larger breweries as they are flexible in their operation<br />
<strong>and</strong> are able to mill malts with differing degrees of modification <strong>and</strong> some adjuncts.<br />
Several arrangements of the rolls <strong>and</strong> screens are used (Sugden et al., 1999; Wilkinson,<br />
2001). In one type malt from the feed roll is crushed by the first pair of rolls <strong>and</strong> then by<br />
the second set of rolls (so at this stage the treatment resembles that given by simple fourroll<br />
mills). The grist from the second pair of rolls falls onto aset of screens, which divide<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1st pair of rolls<br />
Cylindrical<br />
screen with<br />
revolving<br />
beater<br />
Fine grits<br />
<strong>and</strong> flour<br />
Feed roll<br />
Malt<br />
To grist case<br />
Feed valve<br />
Anti-explosion device<br />
Anti-explosion<br />
device<br />
Coarse<br />
grits<br />
2nd pair<br />
of rolls<br />
Fig.5.4 Afour-rollmillusingcylindricalscreenstofractionatethegristfromthefirstpairofrolls<br />
<strong>and</strong> to direct the coarse grits to the second pair of rolls (after Hind, 1950). Similar results are<br />
obtained in mills with oscillating, flat-bed screens.<br />
it into size fractions. Several different arrangements are then possible. For example, the<br />
husk<strong>and</strong>flourfractionsleavethemillwhilethegritsarebrokenmorebypassingbetween<br />
the third pair of rolls. In one particular system the grist from the first <strong>and</strong> second rolls is<br />
divided into two equal streams, which pass to two identical sets of t<strong>and</strong>em flat screens.<br />
This arrangement is compact <strong>and</strong> allows the use of large screen areas (<strong>and</strong> so more<br />
efficient fractionation of the grist) <strong>and</strong>, because the sets of screens are set to oscillate `in<br />
opposition', the vibration of the mill is reduced.<br />
The arrangement of amore common pattern of six-roll mill is shown in Fig. 5.6. Sets<br />
of screens operate between the first <strong>and</strong> second <strong>and</strong> second <strong>and</strong> third pairs of rolls. The<br />
coarse material from the first stage pass to the second set of rolls, the grits go to the third<br />
pair of rolls <strong>and</strong> the flour leaves the mill. The husk <strong>and</strong> flour from the second rolls leave<br />
the mill while the grits pass to the third pair. These mills, appropriately adjusted <strong>and</strong> with<br />
the correctly fluted rollers, can be used to prepare grists that are suitable for use in lauter<br />
tuns or the older types of mash filters. The malt delivered to these mills may be dry or<br />
`conditioned'(see below). Suggested gaps (mm) for the first, second <strong>and</strong> third pairs of<br />
rolls are, (a) for dry malts <strong>and</strong> cereals using alauter tun, 1.6±2.0, 0.7±1. <strong>and</strong> 0.2±0.4; (b)<br />
for conditioned malts using alauter tun, 1.4±1.9, 0.5±1.0 <strong>and</strong> 0.2±0.4, while for dry malt<br />
<strong>and</strong> cereals using an older type of mash filter; (c) the values are 1.0±1.4, 0.4±0.6 <strong>and</strong> 0.1±<br />
0.3 (1mm ˆ0.03937in.). The working rates of six-roll mills are 2±10kg/h/mm roll<br />
length for lauter tun grists <strong>and</strong> 1.5±8kg/h/mm for mash filter grists.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Feed-roll<br />
Pair of<br />
rollers<br />
crushing<br />
the husk<br />
fraction<br />
Second<br />
set of<br />
shaking<br />
screens<br />
5.5 Impact mills<br />
Pair of rollers, carrying<br />
out first crushing<br />
Malt<br />
Husk <strong>and</strong> attached materials<br />
Grits<br />
Flour<br />
Flour Grits Flour<br />
Pair of<br />
rollers<br />
crushing<br />
grits<br />
Husks<br />
with grits<br />
attached<br />
First<br />
set of<br />
shaking<br />
screens<br />
Fig. 5.5 The working principles of a five-roll mill (various sources).<br />
Feed roll<br />
First pair of<br />
crushing rolls<br />
First set of<br />
shaking<br />
screens<br />
Malt<br />
inlet<br />
Flour<br />
Grits Flour<br />
Husks <strong>and</strong> Pair of rollers<br />
attached for reducing<br />
grits coarse grits<br />
Grits retained by<br />
second screen<br />
Coarse materials<br />
retained by top screen<br />
Fig. 5.6 A scheme of a six-roll mill (various sources).<br />
Pair of rollers<br />
crushing<br />
coarse fraction<br />
Second set<br />
of shaking<br />
screens<br />
Disc <strong>and</strong> pin `dry' mills are often used on the experimental or pilot brewery scale <strong>and</strong><br />
they are used in some small breweries (Biche et al., 1999). In these mills the material to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
e ground is delivered to the gap at the centre, between two discs the faces of which are<br />
roughened or covered with short projections. One disc may be static <strong>and</strong> the other rotate<br />
or both may spin, but in opposite directions. The gap between the discs may be adjusted.<br />
As the material moves from the centre to the periphery it is ground. Usually the grist is<br />
carried away in an airflow, but there has been interest in disc milling under water, when<br />
milling <strong>and</strong> mashing in occur together (see Section 5.9).<br />
For years hammer milling has been used to prepare some raw grain adjuncts. It was also<br />
used experimentally <strong>and</strong> in alimited number of breweries for making finely ground grists<br />
from which, after mashing, worts were separated by centrifugation or with rotating vacuum<br />
filters(Briggsetal.,1981;Rehberger<strong>and</strong>Luther,1994).However,theintroductionofhighpressurefilters<strong>and</strong>theMeura2001filterhasencouragedtheuseofhammermillingforgrist<br />
preparation because in these devices the survival of large fragments of husk is irrelevant in<br />
collecting the wort. Hammer mills differ in detail, but the principles of operation are the<br />
same (Fig. 5.7). Malt, sometimes mixed with adjuncts, is fed at apre-determined rate<br />
through arotary valve or afeed roll, into the milling chamber, which is strongly ventilated.<br />
The chamber may be mounted vertically or horizontally. The malt must be scrupulously<br />
cleanedtoremovestones,piecesofmetal<strong>and</strong>`heavyobjects'.Themillingchambercontains<br />
aspinning rotor (e.g. turning at 1500rpm) on which are mounted freely swinging pieces of<br />
metal, the beaters or `hammers', which travel at about 100m/s. The inertia of the rotors is<br />
such that they may take 20±30min. to stop.<br />
Air<br />
inlet<br />
Inlet to<br />
milling<br />
chamber<br />
Hammers<br />
Buffer<br />
hopper<br />
Airflow carrying<br />
grist<br />
Rotary valve/<br />
feed roll<br />
Slide<br />
Rotor<br />
Screen<br />
Anti-vibration<br />
mounting<br />
Fig. 5.7 The layout of atype of hammer mill (various sources).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The impacts of the hammers on the malt smash it up <strong>and</strong> this process continues until the<br />
fragments escape through the semicircular screen that makes up part of the wall of the<br />
chamber. Sometimes the inner wall of the chamber also carries short projections against<br />
which the moving malt can impact. The screens may have mesh widths of 0.5, 1.0mm or<br />
even 2±4mm. The powdered grist is carried out of the mill in the airflow, which transports<br />
it to the grist case, which is equipped with explosion vents. In some mills the rotors are<br />
reversible to obtain even wear on the hammers. These mills are comparatively inexpensive,<br />
but the wear is substantial <strong>and</strong> so maintenance must be regular <strong>and</strong> easy to carry out <strong>and</strong><br />
screens <strong>and</strong> hammers needregular replacement. Such mills mustbe set up onanti-vibration<br />
mountings, <strong>and</strong> housed in acoustically insulated chambers, to muffle the noise.<br />
5.6 Conditioned dry milling<br />
In mashing <strong>and</strong> the subsequent separation of the wort from the spent grains, in mash tuns<br />
or lauter tuns, it is desirable that husk materials should be as nearly intact as possible, but<br />
withdrymilling this is difficulttoachieve while comminuting the endosperm tissue toan<br />
adequate extent. The problem is largely due to the extreme brittleness of the dry husk.<br />
The ideal arrangement for preparing most grists would be to be able to mill malt with<br />
damp, flexible husks but with dry, brittle interiors. To adegree this is achieved by<br />
`conditioning' malt by briefly dampening it, wetting the husk with water or steam, before<br />
it reaches aconventional `dry' mill (Fig. 5.1). The intention is to mill the treated malt<br />
while the husk is damp <strong>and</strong> flexible but before any moisture reaches the endosperm <strong>and</strong><br />
reduces its brittleness. A conditioning screw consists of a screw or paddle conveyor<br />
working in a heated casing. The moving stream of warm malt may be exposed to lowpressure<br />
steam, at 0.5 bar, for 30±60 s. Then, after a 90±120 s equilibration time, the malt<br />
is delivered to a six-roll mill. To dry the equipment <strong>and</strong> minimize corrosion the steam is<br />
turned off 5 min. before the last of the malt passes through. Alternatively, <strong>and</strong> with a<br />
lower risk of enzyme inactivation, water at 30 ëC (86 ëF) or, at least, < 40 ëC (104 ëF), is<br />
sprayed onto the malt <strong>and</strong> then, after a one-minute equilibration period, the malt enters<br />
the mill. The moisture content of the husk is increased by 1.5±1.7%.<br />
The dampened husk is more flexible <strong>and</strong> survives milling better <strong>and</strong> the volume of the<br />
husk sieve fraction is increased by 20±30%. Dry milled grist volumes are 500±700 ml/100 g<br />
while the volumes of conditioned milled grists are 700±1000 ml/100 g. The volumes of the<br />
spent grains are also increased. The mill gaps need to be set closer to obtain the best extract<br />
from conditioned malt. The run-off rate is increased by conditioning <strong>and</strong> lauter tuns may be<br />
loaded more deeply using a mash made with conditioned malt as the bed density is reduced<br />
<strong>and</strong> its porosity is increased (Narziss, 1992; Stoscheck, 1988; Sugden, et al., 1999;<br />
Wilkinson, 2001). It is also said that conditioning gives a better yield of extract, better<br />
attenuation <strong>and</strong> faster saccharification. To prevent clogging dust must be completely<br />
removed from the malt before it reaches the conditioning screw <strong>and</strong> this <strong>and</strong> the adjacent<br />
pieces of equipment must be cleaned regularly.<br />
5.7 Spray steep roller milling<br />
Spray steeping malt before it is milled is a comparatively recent innovation. In spray<br />
steeping mills the malt is held dry in a hopper from which it is fed, at a controlled rate, into a<br />
spray chamber or conditioning shaft or chute, so designed that the malt moves through with a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Mashing liquor<br />
Steep-conditioning<br />
chamber<br />
CIP spray<br />
Sprays for<br />
mashing liquor<br />
Malt inlet<br />
Malt hopper<br />
CIP spray<br />
`plug flow', that is `first in' is `first out' (Fig. 5.8). Here it is sprayed with warm/hot water for<br />
a short time, some quoted conditions being 60±80 ëC (140±176 ëF) for 45±60 s, or 50±70 ëC<br />
(122±158 ëF) for 60±120 s. (Herrmann, et al., 1998; Kunze, 1996; Langenhan, 1992;<br />
Wilkinson, 2001). Surplus water is collected from the spray chamber <strong>and</strong> is re-circulated.<br />
Water usage is 30±80 litres/100 kg malt. Generally the mill has two large, stainless steel rolls<br />
of 290±420 mm (11.41±16.54 in.) diameter with a gap set at 0.2±0.5 mm (0.007±0.0197 in.),<br />
but four-roll versions are also used. If the malt is hard the mill `senses' the increased<br />
workload, so the machine automatically runs more slowly, exposing the malt to a longer<br />
period of wetting. The moisture content of the husk is increased to 18±22%.<br />
Immediately after milling the mashing water is added to the grist <strong>and</strong> the mixture is<br />
pumped to the mash vessel. Lactic acid can be added at this stage to adjust the pH of the<br />
mash. Because this method of milling <strong>and</strong> mashing allows air to be mixed into the mash<br />
there is a chance that unwanted oxidations may occur. This can be prevented by<br />
excluding air, by filling the mill chamber with carbon dioxide or nitrogen gas. Husk<br />
survives milling well, <strong>and</strong> endosperm tissue is adequately broken up. The entire milling/<br />
mashing process is complete in 20±30 minutes. In an alternative arrangement the malt is<br />
CIP<br />
Feed roll<br />
CIP spray<br />
Surplus water<br />
Feed roll<br />
Crushing rolls<br />
Mash<br />
Mash pump<br />
Fig. 5.8 A diagram of a mill using spray-steep conditioning (various sources). CIP, fittings of the<br />
`cleaning in place' system.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
metered into the mill by a rotating, segmented wheel (where it is wetted) which replaces<br />
the conditioning shaft (Kunze, 1996). This system has all the advantages of conditioned<br />
milling, that is, husk survival is enhanced, the grist <strong>and</strong> spent grains occupy a large<br />
volume so their porosity is increased giving shortened lauter times <strong>and</strong> good yields of<br />
extract. Between uses these mills must be thoroughly cleaned. They are equipped with<br />
CIP systems. The performance of these mills is relatively inflexible (Wilkinson, 2003).<br />
5.8 Steep conditioning<br />
In another, older type of wet mill the malt, with or without some raw cereal, is held in a<br />
hopper in which it is steeped (Fig. 5.9; Healy <strong>and</strong> Armitt, 1980; Kunze, 1996; Meisel,<br />
1997; Narziss, 1992; Sugden et al., 1999). The liquor used in the steep is usually at 30±<br />
50 ëC (86±122 ëF) <strong>and</strong> steeping lasts from 10 to 30 minutes. The steep liquor may be<br />
recirculated. After steeping the water is drained away, before the malt is milled.<br />
However, moisture takes time to penetrate the corns <strong>and</strong> so, inside the corns, the first malt<br />
to reach the mill is less moist than the last, so treatment is uneven. The malt has to be<br />
dust-free to prevent clogging. The steeped malt reaches a moisture content of about 30%,<br />
so the contents are partly softened. During milling they are gently squeezed flat by the<br />
large rolls (400 mm, 15.75 in. diameter; 440 rpm, gap 0.30±0.45 mm, approx. 0.012±<br />
0.018 in.) of the mill, squeezing out some of their contents.<br />
A mill may have two or four rolls. As with spray steeping the husks remain intact but<br />
hard ends in the malt are probably not adequately disrupted. Wet milling gives rise to a<br />
large volume of spent grains, indicating that mashes are `open' <strong>and</strong> that the beds have<br />
high porosity. The control of the moisture content of the malt is less exact than that<br />
Spray-head<br />
for steep<br />
Malt inlet<br />
Malt steeping<br />
hopper<br />
Slide<br />
Water Feed roll<br />
To mash<br />
vessel<br />
Steep water<br />
outlet<br />
Drain<br />
CIP ring<br />
Overflow<br />
Crushing rolls<br />
Mashing liquor<br />
Mash-mixing chamber<br />
Homogenizer<br />
Mash pump<br />
Fig. 5.9 A mill using steep conditioning (various sources).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
achieved with spray steeping <strong>and</strong> the mill must be cleaned very well, initially with hot<br />
mashing liquor, to prevent microbes multiplying. The steep water may be discarded, in<br />
which case some extract is lost <strong>and</strong> effluent is produced or it may be added to the<br />
mashing liquor. This may carry aflavour penalty. Mashing liquor is added to the mixture<br />
immediately after milling <strong>and</strong> after mixing the mash is transferred to the mash vessel<br />
using amash pump. Advantages claimed for this system include increased brewhouse<br />
yield, faster wort separation in the lauter tun <strong>and</strong> smoother tasting beers as well as dust<br />
suppression (Stauffer, 1974). On the other h<strong>and</strong> the system is inflexible <strong>and</strong> has other<br />
inherent disadvantages (Wilkinson, 2001). Spray steeping, with its ability to break up dry<br />
<strong>and</strong> brittle endosperm while keeping husk tissue intact, is now preferred.<br />
5.9 Milling under water<br />
Experimentally, other types of wet milling are being investigated. It has been found that<br />
by steeping malt <strong>and</strong> by grinding it under water in adisc mill, afine grist suitable for use<br />
in aMeura 2001 mash filter can be rapidly produced (Biche et al., 1999; De Brackeleire<br />
et al., 2000; Wilkinson, 2003). Milling <strong>and</strong> mashing in occur together. The milling discs<br />
(often 600mm, 23.6in. diameter) rotate at 1275rpm. The gap between the discs is<br />
variable, 0.35±0.55mm (0.0138±0.0217in.) often being used. The smaller the gap<br />
betweentheplatesthefinerthegrindachieved,butwithagreaterpowerconsumption.To<br />
prevent oxidation it is desirable to de-gas the liquor <strong>and</strong> displace the air from the malt<br />
using an inert gas. The system is said to be slightly superior to hammer milling <strong>and</strong> to<br />
give agrist with very good filterability <strong>and</strong> ahigh yield of extract.<br />
Another proposal is to obtain an exceptionally fine grind by breaking up malt <strong>and</strong><br />
adjuncts in a`dispersion chamber' (Menger et al., 2000a,b). In this device the mixture of<br />
malt <strong>and</strong> water passes through aseries of spinning, short, slotted rotors <strong>and</strong> stators which<br />
disrupt it by shear <strong>and</strong> probably impacts. Thus milling <strong>and</strong> mashing in are carried out<br />
together. Again, it is desirable to de-aerate the liquor to minimize oxidative changes.<br />
Within limits it is possible to increase the fineness of the grist by using more disrupting<br />
units in the series, set to achieve increasing degrees of disaggregation.<br />
5.10 Grist cases<br />
With wet mills, milling is coincident with mashing in <strong>and</strong> so the entire operation must be<br />
completed in acomparatively short time, say 20±30 minutes. However, this is not the<br />
case with dry milling as the grist can be accumulated <strong>and</strong> stored, at least for ashort<br />
period, ready for mashing in, in aseparate operation. Dry grists can be evaluated using<br />
st<strong>and</strong>ard sieves (Tables 5.1, 5.4). Grists can have low bulk densities; for example, 100kg<br />
can occupy 3 hl (approx 20.8 lb./ft. 3 ). Thus a dry milled grist may occupy 2.6 hl/kg, while<br />
a conditioned malt grist might occupy 3.2 hl/ kg (Narziss, 1992). Thus grist cases, the<br />
containers in which grists are stored, have large volumes <strong>and</strong> must contain enough grist<br />
for a mash. Usually they are made of mild steel, <strong>and</strong> are designed to contain dust <strong>and</strong><br />
exclude steam <strong>and</strong> damp air, while being able to release their load under gravity, at the<br />
required rate. Sometimes the cases are fed from a mill <strong>and</strong>, by a proportional feeder, from<br />
a store of an adjunct such as wheat flour, that needs to be mixed uniformly into the grist.<br />
The grist may be directed to the mashing in system through a chute or by way of a gentle<br />
belt conveyor. Unless the grist is intended for a new type of mash filter abrasion caused<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
y rough conveying or h<strong>and</strong>ling should be avoided so that the grist is not broken down<br />
further. All the components of a grist should be well mixed, <strong>and</strong> so any vibrations or<br />
conveying that favour the segregation of the components of the grist, for example, into<br />
layers of husk <strong>and</strong> fines, must be avoided. Commonly a grist case is above one mashing<br />
vessel, which it serves. However, sometimes a grist case can be equipped with chutes or<br />
conveyors that allow it to serve more than one mashing vessel, while other grist cases can<br />
be moved so that they can serve more than one vessel <strong>and</strong> several containers, each<br />
containing enough grist for one mash, may be grouped together in one structure.<br />
In two old processes, that have fallen out of use, the grist in the case could be heated by<br />
steam-heated pipes placed in the grist case. In `retorrification' the grist was warmed to the<br />
mashing-in temperature. Thus the temperature of the mashing liquor did not need to be<br />
greatly above that of the grist to achieve the correct initial heat, <strong>and</strong> so thermal inactivation<br />
of enzymes, caused by local overheating, was minimized on mashing in <strong>and</strong> more uniform<br />
mashing temperatures were achieved when the ambient temperature fluctuated.<br />
When grists were subjected to `aromatization' they were heated to about 130 ëC<br />
(266 ëF) for 15 minutes to increase the colour <strong>and</strong> aroma. Inevitably this must have caused<br />
some enzyme inactivation.<br />
5.11 References<br />
BHADURI, R. (1996) Brew. Distill. Internat., Mar., 18.<br />
BICHE, J., HARMEGNIES, F. <strong>and</strong> TIGEL, R. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes, p. 593.<br />
BRIGGS, D. E. (1998) Malts <strong>and</strong> Malting. London. Blackie Academic <strong>and</strong> Professional, 796 pp.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting <strong>and</strong> <strong>Brewing</strong> Science (2nd edn).<br />
Vol. I. Malt <strong>and</strong> Sweet Wort. London. Chapman <strong>and</strong> Hall, 387 pp.<br />
CRESCENZI, A. M. (1987) J. Inst. <strong>Brewing</strong>, 93, 193.<br />
DE BRACKELEIRE, C., HARMEGNIES, F., TIGEL, R. <strong>and</strong> MENDES, J. P. (2000) Brauwelt Internat., 18 (5), 372.<br />
FARNISH, R. (2002) Brewers' Guard., Feb., p. 26.<br />
HEALY, P. <strong>and</strong> ARMITT, J. D. G. (1980) Proc. 16th Conv. Inst <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong><br />
Section), Sydney, p. 91.<br />
HERRMANN, H., KANTELBERG, B. <strong>and</strong> WIESNER, R. (1998) Brauwelt Internat., 16 (1), 44.<br />
HIND, H. L. (1950) <strong>Brewing</strong> Science <strong>and</strong> Practice, Vol. II. <strong>Brewing</strong> Processes. London, Chapman <strong>and</strong> Hall,<br />
pp. 507±1020.<br />
ISOE, A., KANAGAWA, K., ONO, M., NAKATANI, K. <strong>and</strong> NISHIGAKI, M. (1991) Proc. 23rd Congr. Eur. Brew.<br />
Conv., Lisbon, p. 697.<br />
KROTTENTHALER, M., ZUÈ RCHER, J., SCHNEIDER, J., BACK, W. <strong>and</strong> WEISSER, H. (1999) Proc. 27th Congr.<br />
Eur. Brew. Conv., Cannes, p. 603.<br />
KUNZE, W. (1996) Technology <strong>Brewing</strong> <strong>and</strong> Malting. (Wainwright, T. Transl.) Berlin, VLB, 726 pp.<br />
LANGENHAN, R. (1992) Brew. Distill. Internat., 23 (5), 16.<br />
MEISEL, D. (1997) A Practical Guide to Good Lager <strong>Brewing</strong> Practice. Inst. <strong>Brewing</strong>, (Central <strong>and</strong><br />
Southern African Sect.).<br />
MENGER, H.-J., SALZGEBER, G. <strong>and</strong> PIEPER, H. J. (2000a) Brauwelt Internat., 18 (1), 54.<br />
MENGER, H.-J., MIROLL, F., FORCH, M., BIURRUN, R., SCHWILL-MIEDANER, A., HERRMAN, J. <strong>and</strong> RAPP, T.<br />
(2000b) Brauwelt Internat., 18 (2), 120.<br />
NARZISS, L. (1992) Die Bierbrauerei.(7nt Auflage) Bd. II. Die Technologie der WuÈrzebereitung. Stuttgart.<br />
Ferdin<strong>and</strong> Enke Verlag, 402 pp.<br />
NICOL, S. O. <strong>and</strong> ANDREWS, J. M. H. (1996) Ferment, 9 (3), 145.<br />
REHBERGER, A. J. <strong>and</strong> LUTHER, G. E. (1994) in H<strong>and</strong>book of <strong>Brewing</strong> (Hardwick, W. A. ed.). New York,<br />
Marcel Dekker Inc., p. 247.<br />
STAUFFER, J. (1974) MBAA Tech. Quart., 11, (4), 240.<br />
STOSCHECK, W. (1988) MBAA Tech. Quart., 25 (2), 108.<br />
STUBITS, M., TENG, J. <strong>and</strong> PEREIRA, J. (1986) J. Amer. Soc. Brew. Chem., 44 (1), 12.<br />
SUGDEN, T. D., WEBB, C., BYRNE, H., VAN WAESBERGHE, J. <strong>and</strong> WULFF, T. (1999) Milling. E. B. C.<br />
H<strong>and</strong>book of Good Practice. NuÈrnberg. Hans Carl, 102 pp.<br />
WILKINSON, R. (2001) Brewers' Guard., 130, (4), 29.<br />
WILKINSON, R. (2003) Brewers' Guard., 132, (1), 26.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
6<br />
Mashing technology<br />
6.1 Introduction<br />
Mashing is the process by which sweet wort is prepared. It involves `mashing in', the<br />
mixing of the milled grist <strong>and</strong> the brewing liquor at the correct temperature <strong>and</strong> in the<br />
correct proportions to obtain the mash. After aperiod, with or without temperature<br />
changes, during which the necessary biochemical changes occur, the liquid `sweet wort',<br />
which contains the extract, is separated from the residual solids, the `spent grains' or<br />
`draff'. Some extract remains in the draff, <strong>and</strong> as much of this as possible is recovered by<br />
`sparging', washing the grains with hot brewing liquor.<br />
In traditional brewing, as practised in homes or small inns, hot water was placed in a<br />
wooden tub or tun, <strong>and</strong> the grist (malt that had been ground between millstones) was<br />
mixed <strong>and</strong> mashed in by stirring or rowing with arake, paddle or `oar' (Fig. 6.1). No<br />
reliable means of measuring temperatures was available. In one method, which gave rise<br />
to the `classical' British infusion system, the water temperature was guessed to be<br />
suitable by feel or by how clearly the brewer's face was reflected in the water. After a<br />
period a basket was pushed into the mash <strong>and</strong> wort that seeped into it was ladled into a<br />
receiver, in readiness for boiling with hops or other flavouring herbs. When wort<br />
recovery became difficult more hot water was mixed into the mash (re-mashing) <strong>and</strong><br />
another, weaker wort was recovered. This was repeated until the worts were too weak to<br />
be worth collecting. In later times wort was collected from mashes using primitive mash<br />
tuns, in which the wort drained from the mash through a perforated `strainer' in the base<br />
of the tun. The structures of old (approx. 200 years), relatively small British country<br />
house breweries are documented (Sambrook, 1996).<br />
In an alternative method, which gave rise to the classical mainl<strong>and</strong> European<br />
decoction mashing system, traditionally used for brewing lager beers, the mash was made<br />
with slightly warm water. At intervals a `decoction' was carried out, that is, a proportion<br />
of the mash, perhaps one-third, was withdrawn <strong>and</strong> slowly raised to boiling in the copper<br />
that would later be used for boiling the wort. The hot mash was then transferred back to<br />
the `main mash', <strong>and</strong> was mixed in. In this way the temperature of the whole mash was<br />
increased. Repeated decoctions increased the mash temperature in steps, an approach that<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
minimized the risk of overheating <strong>and</strong> premature total enzyme destruction. In more<br />
modern variations of this system the temperatures achieved (now exactly controlled) are<br />
optimal for various enzyme-catalysed processes in the mash <strong>and</strong> allow comparatively<br />
under-modified malts to be mashed successfully (Chapter 4). Raw cereal adjuncts may<br />
conveniently be cooked in decoction vessels. In contrast, the traditional infusion system<br />
of mashing requires well-modified malts <strong>and</strong> not more than about 20% of unmalted<br />
adjuncts (Chapter 4).<br />
The equipment used in breweries has been progressively refined <strong>and</strong> there has been a<br />
convergence in the <strong>practice</strong>s of ale <strong>and</strong> lager brewers. The motives for these alterations<br />
are chiefly economic. It is often desirable to maximize productivity (`throughput', the<br />
numberofbrewscompletedevery24h)<strong>and</strong>torecoverasmuchextractasiseconomically<br />
worthwhile from agiven grist (Chapter 18). It is also necessary to reproducibly recover a<br />
certain volume of wort having exactly the characteristics needed to make aparticular<br />
beer. At the same time energy <strong>and</strong> water usage must be minimized <strong>and</strong> so must the<br />
production of effluents. At the present time there are breweries operating with many<br />
different kinds of `traditional' <strong>and</strong> `modern' equipment. The more common types will be<br />
described. Sometimes old types of plant have been retained, despite some inconvenience<br />
or poorer efficiency, because a newer system has not been able to produce abeer<br />
matching that produced by the old system. In the case of smaller breweries older kinds of<br />
equipment may be retained because of its simplicity, or because replacement is not<br />
economic. While older equipment is often made of attractive polished wood <strong>and</strong> copper,<br />
in newer equipment these materials have been largely replaced by the cheaper, more<br />
deterioration-resistant stainless steels (Chapter 10).<br />
6.2 Mashing in<br />
Fig. 6.1 Awooden mashing rake or oar.<br />
Mashing in, the process of mixing the mashing liquor <strong>and</strong> the grist, is critical. The<br />
proportions of liquor to grist <strong>and</strong> the temperature of the mixture must both be correct <strong>and</strong><br />
the grist must be evenly mixed in the mash, with no clumping or `balling', which reduces<br />
extract recovery, <strong>and</strong> no segregation of the grist components. In traditional infusion<br />
mashing,whereitisnoteasytoadjustthemashtemperature,thevalueinitiallyattainedis<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Slide valves<br />
Hot liquor<br />
Drive<br />
Grist case<br />
‘Worm’ or<br />
screw<br />
Beater<br />
rods<br />
Swivel mount<br />
To mash<br />
Fig. 6.2 Steel's mashing machine, used in traditional British infusion mashing. In more modern<br />
versions the rate of grist flow is regulated mechanically <strong>and</strong> the slide valves prevent steam from the<br />
mash travelling into the grist case.<br />
critical. Historically, grist was poured into attemperated water in the mash tun or `tub'<br />
<strong>and</strong> was mixed in by `rowing', alaborious process that often resulted in inadequate<br />
mixing <strong>and</strong> balls of unmixed grist. Mechanical means of mixing mashes were necessary.<br />
Mash tun rakes were invented by Matterface in 1807 (see below; Sykes <strong>and</strong> Ling, 1907).<br />
This kind of equipment remained widely in use in the UK until the 1960s. It is now rare.<br />
Several `external' mashing machines were invented <strong>and</strong>, in various forms, some of these<br />
are still in use.<br />
The Steel's masher, introduced in 1853, consists of ahorizontal tube about 46cm<br />
(18in.) in diameter <strong>and</strong> was once typical of ale breweries (Fig. 6.2). Grist from the gristcaseisdeliveredtooneendviaaverticaltubeatacontrolledrate.Slidevalvesareusedto<br />
prevent vapour from the mashrising intothe grist case.Thegrist issprayedwith mashing<br />
liquor <strong>and</strong> the wetted material is driven along the tube by ascrew conveyor <strong>and</strong> is then<br />
mixed by aseries of short beater-rods mounted on the same shaft as the conveyor screw.<br />
The mashed material is then dropped from aspout into the vessel below. This device is<br />
capableofmixingthethickmashestypical oftraditionalinfusionmashing(e.g.1.6±3.2hl<br />
liquor/100kg grist; 9.9±19.9imp. brl/ton). With this system it is impossible to prevent<br />
oxygen uptake. Sometimes the masher is mounted so that it can be swivelled sideways<br />
<strong>and</strong> so can deliver into either of two vessels.<br />
Mash hydrators, or `pre-mashers', are designed for making the thinner mashes used in<br />
decoction mashing or temperature-programmed infusion mashing (3.3±5hl liquor/100kg<br />
grist; 20.5±31.0imp. brl/ton), which must be stirred <strong>and</strong> pumped between vessels. In<br />
general, light beers are made with more dilute mashes than those used for dark beers. In<br />
mash hydrators grist, falling down a tube at a controlled rate, meets a spray of<br />
attemperated water flowing at acontrolled rate from aperforated central tube or from a<br />
surrounding casing (Figs 6.3, 6.4). In other increasingly common devices the water is<br />
injected tangentially creating avortex, which rapidly mixes with the grist (Fig. 6.5). This<br />
device contains dust <strong>and</strong> minimizes oxygen pick up. It is desirable that the mash from the<br />
hydratordoesnotfallintothemash,butisdirected ontothesideofthemashingvessel, or<br />
is pumped into the base, so that it slides down to the mash with the minimum uptake of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Thermometer<br />
Liquor<br />
inlet<br />
Slide<br />
oxygen (Wilkinson, 2003). None of these devices has moving parts to wear out. Asprayball<br />
should always be mounted above the liquor inlet to facilitate `cleaning in-place',<br />
CIP. The grist from the hydrator is not uniformly hydrated. Hydration is completed<br />
during stirring in the mashing vessel.<br />
In addition to the methods mentioned, wet milling coincides with mashing in <strong>and</strong> so<br />
with this approach no grist case is involved (Chapter 5). In some plants the material from<br />
the pre-masher enters an inclined disc mixing vessel. This consists of a cylinder with<br />
rounded ends mounted on its side <strong>and</strong> having a longitudinal shaft mounted about a third<br />
Grist<br />
Window<br />
To mashing vessel<br />
Fig. 6.3 Premasher or Maitl<strong>and</strong> grist hydrator, of a traditional mainl<strong>and</strong> European type, used in<br />
conjunction with decoction mashing (after Narziss, 1992). Note that the grist falls past a perforated<br />
tube, which delivers the mashing liquor. The mixture pushes open a counter-balanced door <strong>and</strong> falls<br />
into the vessel.<br />
Grist<br />
Mash<br />
Water<br />
Fig. 6.4 An alternative type of mash hydrator (after Rehberger <strong>and</strong> Luther, 1994).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Grist<br />
Mash<br />
of the way up from the bottom. The shaft carries aseries of discs inclined at angles to the<br />
shaft which rotates. The discs mix the mash very gently <strong>and</strong> thoroughly (Kunze, 1996).<br />
In most plants it is usual for the pre-hydrated grist to be transferred into amashing<br />
vessel, where mixing occurs. To minimize oxidative changes in the mash or wort, the<br />
mash may be gently pumped into the vessel at the side or up through the base, so<br />
minimizing turbulence <strong>and</strong> the uptake of oxygen. Degassing the mashing liquor, flushing<br />
the grist with an inert gas such as nitrogen or carbon dioxide <strong>and</strong> filling the base of the<br />
vessel with an inert gas to displace air <strong>and</strong> so limit oxygen pick-up may also be used<br />
(Yamaguchi et al., 1997).<br />
In many breweries the older names for critical temperatures are retained (Hind, 1940).<br />
Thus liquor heat <strong>and</strong> striking heat are terms for the temperature of the mashing liquor in<br />
the hot water tank <strong>and</strong> at the mashing machine respectively. The initial heat is the<br />
temperature of the freshly mixed mash, sparge heat is the temperature of the sparging<br />
liquor <strong>and</strong> the tap heat is the temperature of the wort as it is drawn off. In infusion<br />
mashing in modern mash tuns the opportunities for adjusting the mash temperature are<br />
limited <strong>and</strong> it is essential to achieve the correct initial heat (temperature) when mashing<br />
in. In modern plant this may be achieved automatically by varying the temperature of the<br />
mashing liquor sothat the correct initial heat isattained. In older plant this isachieved by<br />
askilled operator's judgement, guided initially by calculation <strong>and</strong> later by experience.<br />
When malt is mixed with water heat is generated <strong>and</strong> this slaking heat or heat of<br />
hydration is less for malts with higher moisture contents (Table 6.1). So, for example, a<br />
malt with amoisture content of 2%, mashed in aparticular way, may give atemperature<br />
rise of about 4.8ëC (8.6ëF), while amalt with amoisture content of 6% would give a<br />
temperature rise of 2.6ëC (4.6ëF) when mashed in the same way. The initial heat of the<br />
mash can be calculated from the formula<br />
I=[St +RT/S+R] +[0.5H/S+R]<br />
Water<br />
Fig. 6.5 Avortex mash mixer (after Rehberger <strong>and</strong> Luther, 1994). The grist falls into aswirling<br />
tube of liquor from that tangentially injected into the outer chamber. With these, <strong>and</strong> other prehydrators,<br />
it is desirable to have a mixing chamber through which the grist must pass before it<br />
leaves the unit.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 6.1 The specific heats <strong>and</strong> slaking heats of malt at different moisture contents. Other malts<br />
may have slightly different values. Data recalculated, by interpolation, from the data of (a) Brown<br />
(1910) <strong>and</strong> (b) Hopkins <strong>and</strong> Carter (1933)<br />
Moisture Specific Slaking heat at 65 ëC (150 ëF)<br />
(%) heat (a) g.cal/ëF g.cal/ëC<br />
where Sˆspecific heat of the malt, tˆthe temperature of the malt, Rˆthe weight of<br />
water, relative to the unit weight of the malt, Tis the temperature of the water, His the<br />
slaking heat of the malt expressed in the correct units <strong>and</strong> Iis the initial temperature of<br />
the mash. Others prefer to use H(rather than 0.5H) <strong>and</strong> make allowances for heat losses,<br />
determined by trial <strong>and</strong> error. The final temperature of amash warmed by `underletting',<br />
that is the addition of hot liquor to the mash, can be calculated from the formula<br />
Final temperature =[M(S+R) +QT] /(S+R+Q)<br />
where Mˆtemperature of the mash at the time of underletting, Qis the quantity of water<br />
used in the underlet, Tˆtemperature of the underlet liquor. The other symbols are as<br />
used before. These calculations can be used only for guidance. They cannot give exact<br />
results because no allowance is made for heat losses from the system, <strong>and</strong> these will vary<br />
with the temperature of the brewhouse.<br />
6.3 The mash tun<br />
(a) (b) (a) (b)<br />
0 0.38 28.0 33.4 15.6 18.6<br />
1 0.38 24.7 28.7 13.7 15.9<br />
2 0.39 21.5 24.9 11.9 13.8<br />
4 0.40 15.8 18.8 8.8 10.4<br />
6 0.41 11.5 14.2 6.4 7.9<br />
8 0.42 8.5 11.7 4.7 6.5<br />
The mash tun (`kieve' in Irel<strong>and</strong>) is the simplest device for mashing <strong>and</strong> preparing sweet<br />
wort.Mashconversion<strong>and</strong>wortseparationfromthespentgrainstakeplaceinonevessel,<br />
<strong>and</strong> so mash tuns should be distinguished from the mash mixing <strong>and</strong> incubation vessels<br />
which are used to carry out the mash conversion step only, wort collection being carried<br />
out in aseparate device, usually alauter tun or amash filter.<br />
6.3.1 Construction<br />
Mash tuns are circular in cross-section <strong>and</strong> vary greatly in size, but they are usually 2.0±<br />
2.5m(approx. 6±8ft.) in depth. Originally the tuns were of wood <strong>and</strong>, for many years<br />
mostwereequipped withrakingmachinery(Fig.6.6). Wood ishard toclean<strong>and</strong>hasonly<br />
alimited life. Increasingly, mash tuns were made of iron or copper or, more recently,<br />
stainless steel. The properties of some materials used in the construction of brewing<br />
vesselsareoutlinedinChapter10.Theyareinsulatedatthebase<strong>and</strong>aroundthesides<strong>and</strong><br />
areoftencladwithwood.Originallythetunswereopen,butnowtheyareusuallycovered<br />
with metal domes, equipped with inspection ports <strong>and</strong> lights.<br />
Covering atun slows heat loss <strong>and</strong> prevents the water vapour from the mash spreading<br />
through the brewing room (Fig. 6.7). Sometimes the vapours are carried away in apipe<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Rake<br />
Axis, carrying rakes,<br />
which rotates <strong>and</strong><br />
moves around the tun<br />
False, gun-metal<br />
plates<br />
True base of<br />
tun (wood)<br />
Spent grains<br />
discharge<br />
Rotating<br />
sparge arm<br />
Driving<br />
machinery<br />
Hot<br />
liquor<br />
Vertical<br />
rotating drive<br />
Wooden<br />
wall of tun<br />
Fig. 6.6 An old pattern of mash tun as used in about 1880 (various sources). Some few of this<br />
general type are still in use.<br />
fitted to the dome. The true base of the tun is covered by a`false bottom' or deck. This<br />
consists of interlocking slotted (or drilled in some small tuns) metal plates of gun-metal<br />
or stainless steel or of stainless steel wedge wire (Figs 6.8, 6.9) mounted on short legs<br />
about 5±7.6cm (2±3in.) above the true base. The plates interlock in aunique pattern <strong>and</strong><br />
can be lifted for cleaning or repairs. The slots give afree area of about 10±12% of the<br />
false bottom, while wedge wire gives the same or ahigher value (up to 22%). The slots<br />
are typically 0.7±1.0mm (0.028±0.039in.) wide at the top <strong>and</strong> widen out below to<br />
facilitate cleaning <strong>and</strong> reduce the chances of fragments of grist wedging in the spaces. In<br />
the past all cleaning was by h<strong>and</strong>, involving the regular lifting of the plates of the false<br />
bottom <strong>and</strong> later re-assembling them, but now sprays may be fitted to allow CIP both<br />
above <strong>and</strong> below the deck. One, or more, large holes, fitted with valves, are in the false<br />
bottom. These are opened at the end of mashing to allow the removal of the spent grains<br />
<strong>and</strong> some rinse water.<br />
In the past, grain removal was by h<strong>and</strong> but now, except in small breweries where<br />
manual removal is still used, grains are swept out of the discharge ports by horizontally<br />
rotating arms. The grains are collected <strong>and</strong> transferred by screw conveyor or compressed<br />
air to the collection silo for removal to farms or animal compounders. An alternative<br />
procedure of slurrying the grains with water <strong>and</strong> pumping the mixture toacollection tank<br />
has been discontinued because wetting the spent grains reduces their value <strong>and</strong> the use of<br />
more water <strong>and</strong> the production of effluent with ahigh BOD increases costs. Few English<br />
breweries have retained rotating rakes. The machinery is clumsy <strong>and</strong> it is suspected of<br />
allowing channelling during sparging, with aconsequent reduction in extract recovery.<br />
On the other h<strong>and</strong> rakes facilitate mixing the mash with underlet water, allowing<br />
controlled increases in temperature.<br />
Some few mash tuns have been equipped with knives for `cutting' the mash <strong>and</strong> for<br />
discharging the grains, as is done in many lauter tuns (Section 6.5). Originally drained<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Device for<br />
adjusting<br />
hydrostatic<br />
head<br />
Wort to<br />
underback<br />
Grist case<br />
Steel’s masher<br />
Run-off pipe<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Cork<br />
insulation<br />
on true base<br />
Hot liquor<br />
to sparge Bearings for<br />
sparge arm<br />
Motor for<br />
driving grains<br />
discharge gear<br />
Sliding inspection<br />
hatches<br />
Fig. 6.7 A current pattern of mash tun (after Briggs et al., 1981).<br />
Grains<br />
discharge<br />
pipe<br />
Sparge arm<br />
Wood <strong>and</strong> cork<br />
insulation<br />
Mash<br />
Floor<br />
Grains<br />
discharge<br />
arm<br />
False bottom
1 in. 1 cm Scale<br />
Fig.6.8 Slotsinplatesthatformthefalsebottomsofmashtuns(afterBriggsetal.,1981).(a)Plan<br />
view of the upper surface. (b) Vertical section across the slots. (c) Plan view of the lower surface.<br />
(d) Avertical section, at right-angles to (b), along two slots.<br />
Fig. 6.9 Apiece of wedge wire (after various sources).<br />
grains were remashed withfreshhot watertorecoverentrainedextract. Now`sparging' is<br />
employed. Hot water is sprinkled onto the surface of the mash from centrally mounted,<br />
rotating perforated tubes, the sparge arms. These may be mechanically driven, but more<br />
usually they are driven by their reaction to the streams of water coming from the<br />
perforations, which are at an angle to the vertical (Fig. 6.7). Because it is desirable to<br />
apply the sparge liquor equally to all areas of the mash the perforations are more widely<br />
spaced towards the centre of the tun <strong>and</strong> more closely spaced towards the periphery.<br />
During wort collection the liquid is driven down through the bed of grain by the pressure<br />
difference between the liquid at the top of the mash <strong>and</strong> the pressure below the false<br />
bottom. This pressure difference is measured. To control this pressure difference, <strong>and</strong><br />
hence the `suction' on the grain bed <strong>and</strong> its compression, various devices, which act as<br />
weirs, may be used including aswinging, inverted U-tube (a Valentine tube), or adevice<br />
equivalent to it (Fig. 6.7). Alternatively the rate of collection can be regulated with a<br />
pump, <strong>and</strong> the rate of sparging can be set to equal the rate of run-off using two matched<br />
pumps. Excessive pressure can force the goods down onto the plates <strong>and</strong> cause blockages<br />
or so compress the bed that the resistance to the flow of the liquid through the bed is high<br />
<strong>and</strong> run-off is slowed. The result is aset mash.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
(a)<br />
(b)<br />
(c)<br />
(d)
To obtain an even run-off <strong>and</strong> sparge it is desirable to have multiple wort collection<br />
pipes,connectedtothetruebottomofthetun,distributedinanevenpattern.Theobjectis<br />
to have each pipe draining the same area beneath the deck, say 1.9±2.3m 2 (20±25ft. 2 ).In<br />
the traditional arrangement these discharge the wort via taps <strong>and</strong> swan-necked tubes into<br />
an open collecting trough. This allows the clarity, the gravity <strong>and</strong> the rate of flow of the<br />
wort from each tap to be checked. By manipulating the taps the rate of flow can be<br />
regulated <strong>and</strong> balanced. Differences in the quality of the wort from different taps<br />
indicates problems with the quality of the grain bed or the sparge.<br />
6.3.2 Mash tun operations<br />
Before use the clean mash tun is preheated to the mashing temperature with steam or hot<br />
water.Alltapsareclosed, thetemperatureofthemashingliquor ischecked <strong>and</strong>soarethe<br />
contents of the grist case. Some liquor is admitted into the base of the tun to drive air out<br />
from beneath the false bottom <strong>and</strong> to cover the plates with alayer of water 2±5cm<br />
(roughly1±2in.)deep.ThiscushionsthemashasitisdroppedinfromtheSteel'sorother<br />
premasher, helps the mash to spread across the false base <strong>and</strong> reduces the chance of the<br />
slots being blocked. Mashing in may take 20±30min. In modern mash tuns the mash<br />
entrains air <strong>and</strong> most of the goods float (Harris, 1968, 1971). In older tuns the mash was<br />
raked for afew minutes to mix the contents. The depth of the mash is typically 0.91±<br />
1.5m(approx. 3±5ft.), but depths of up to 2.74m(9ft.) have been used. One advantage<br />
of mash tuns is that they can be used with mashes of widely varying volumes.<br />
The mash then begins its st<strong>and</strong>, normally aperiod of 1.25±2.5h, to allow all necessary<br />
biochemical processes to take place. The temperature of the mash during the st<strong>and</strong> is not<br />
normally altered, but this is acomparatively recent development. Formerly moderate<br />
upward temperature adjustments were usual (Chapter 4; Hind, 1940; Sykes <strong>and</strong> Ling,<br />
1907).Theapplicationofheattoathickmashthroughaheaterinthewalls<strong>and</strong>baseofan<br />
unstirredmashtunisnotefficient.Atvarioustimesheatincreaseswerebroughtaboutwith<br />
steam-heated coils beneath the false bottom or by direct steam injection, by cycling wort<br />
fromthebaseofthetuntothetopofthemashviaanexternalheater,bya`steamplough'(a<br />
device heated by steam or water which rotated with the machinery near to the base of the<br />
tun) or by underletting (sometimes with simultaneous sparging). With underletting, the<br />
mostusualmethod,thecalculatedvolumeofhotliquor,atanappropriatetemperature,was<br />
slowly let into the base of the tun, beneath the plates <strong>and</strong> raising the mash bed without<br />
disruptingit,thenthemashwasmixed,soitstemperaturewasincreased<strong>and</strong>itwasdiluted.<br />
Sometimes wort run off becomes slow or even ceases. This is liable to occur if poor<br />
quality malt has been used, if the grist has been milled too finely, if the grist contains a<br />
high proportion of adjuncts, or if the operator has tried to draw off wort too quickly <strong>and</strong><br />
has pulled the mash down onto the plates <strong>and</strong> caused the bed to become compressed.<br />
Such a`set mash' may be cleared by using an underlet to lift the goods off the plates. Set<br />
mashes caused by poor quality grist components are made less likely by adding an<br />
enzyme preparation containing -glucanase, cellulase <strong>and</strong> pentosanase activities.<br />
Drainage from the mash is dependent on it having an open structure <strong>and</strong> this, in turn,<br />
requires that the malt husk fraction be damaged as little as possible during milling<br />
(Chapter 5). In the past some brewers used aproportion of `husky' malt or even added<br />
chopped straw or oat husks to `open-out' the mash <strong>and</strong> aid drainage. The addition of such<br />
materials is undesirable as unwanted flavours can be conferred to the beer.<br />
When the st<strong>and</strong>, with or without an underlet, is completed the first wort is run off. The<br />
first runnings may be a little turbid, <strong>and</strong> these are usually pumped back to the top of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mash<strong>and</strong>recirculated.Whenthewortisperfectlyclear(i.e.is`runningbright')theflowis<br />
diverted <strong>and</strong> the liquid is either collected in aholding vessel, an underback, where it is<br />
maintained at 71±82ëC (160±180ëF) to prevent the multiplication of contaminating<br />
microbes, or it is transferred directly to the copper. When the `first worts' have been<br />
collected <strong>and</strong> the mash has settled, but before the surface of the grains has become dry,<br />
sparging begins, the hot water (usually at about 78ëC, 172.4ëF) being sprinkled on from<br />
therotatingspargearms.Thespargingratematchestherateofwortrunoff,<strong>and</strong>theliquor<br />
displaces the wort downward <strong>and</strong> through the plates <strong>and</strong> leaches extract from the grist<br />
particles.Thehightemperaturefacilitatesextractrecoverybecauseitreducestheviscosity<br />
ofthewort<strong>and</strong>sofacilitatesrunoff<strong>and</strong>itacceleratestheleachingofextractfromthegrist.<br />
Afterthefirstwortshavebeencollectedthespecificgravityofthecollectedwortbegins<br />
to fall. The decline continues until achosen gravity (often SG 1003±1005, 0.78±1.3ëP) is<br />
reached,whencollectionceases.Thegoodsarethenallowedtodrain.Thedrainingsmaybe<br />
senttowaste,withaconsequentlossofextract<strong>and</strong>atthecostofaneffluentcharge,orthey<br />
may be stored hot for ashort period <strong>and</strong> then be mixed into asubsequent mash, when the<br />
extractisrecovered.However,thequalityoftheextractinthedrainingsisinferiortothatin<br />
themainwort.Spargingoftenlasts 4±5h.The sparge liquorshouldnotcontain `temporary<br />
hardness' <strong>and</strong> should contain an adequate level of calcium ions (Chapter 4). At higher pH<br />
values more tannins <strong>and</strong> undesirable materials are extracted, leading to reductions in beer<br />
quality. About 5.35hl of sparge liquor may be used for 100kg malt (33imp. brl/ton) so in<br />
allatotalofabout8.1hlofliquorareusedforeach100kgmalt(50.3imp.brl/ton).Someof<br />
this liquor, perhaps 2%, remains in the spent grains. The spent grains have amoisture<br />
content of about 80%. These are discharged <strong>and</strong> the tun is cleaned before the next mash.<br />
The entire cycle time with amash tun is often 5±9h, but in one famous brewery the time<br />
was 18h<strong>and</strong> mashing in began at midnight. So mash tuns generate clear, high-quality ale<br />
worts which, because the mashes are thick, can produce high-gravity worts but by modern<br />
st<strong>and</strong>ards their turn-round times are slow. Oxygen exclusion is not feasible but for<br />
traditional ales it is not necessary. The thick bed of grains allows the production of very<br />
brightwortsbutalsoensuresthattherunofftimeiscomparativelylong.Extractrecoveries<br />
(relative to the laboratory extract) of 98% have been claimed, <strong>and</strong> 96±97% is usual but in<br />
some small breweries the recovery may be as little as 85%. Other disadvantages of mash<br />
tuns are their inflexibility with regard to mashing temperatures, their requirement for a<br />
coarsely ground grist (with aconsequent reduction in extract recovery), the need for well<br />
modified malts <strong>and</strong> the difficulty in using wet, cooked adjuncts.<br />
6.4 Mashing vessels for decoction, double mashing <strong>and</strong><br />
temperature-programmed infusion mashing systems<br />
In these mashing systems mashes are relatively thin (3.3±5hl liquor/100kg grist; 20.48±<br />
31imp. brl/ton) <strong>and</strong> so they can be stirred <strong>and</strong> pumped between vessels. While, in<br />
principle, the types of vessels have remained constant for aconsiderable time there have<br />
been significant improvements in design <strong>and</strong> in the designs of the wort separation<br />
devices, the lauter tuns <strong>and</strong> mash filters, used in conjunction with them.<br />
6.4.1 Decoction <strong>and</strong> double mashing<br />
Inthetraditionaldecoctionmashingsystemgristismashedinusingahydrator<strong>and</strong>fallsinto<br />
amash-mixingvessel(Fig.6.10).Tominimizeoxygenpick-upthemashmayberundown<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Registering<br />
thermometer<br />
Propeller<br />
Worm<br />
gear<br />
Throttle<br />
valve<br />
Mashing<br />
valve<br />
Pipe for<br />
ground malt<br />
Pre-masher<br />
Fig. 6.10 A section of an older pattern of a decoction mash-mixing vessel (after Hind, 1940).<br />
thesideofthevesselorpumpedintothebase.AtthisstagepHadjustmentmaybeachieved<br />
bytheadditionofafood-grademineralacidorbiologically preparedlacticacid.Acidmalt<br />
may have been included in the grist to achieve the same effect. Mashes are stirred, usually<br />
by apropeller in the base of the circular vessel. However, vessels with rectangular cross<br />
sections have been used, the contents being mixed by stirrers mounted on long shafts<br />
extending downwards from above. The older vessels are often of copper, have adouble<br />
walled steam heating jacket <strong>and</strong> are well insulated. Heating may also be by internal steam<br />
heated coils or direct steam injection. Low-pressure steam (15 bar, 171ëC, 340ëF) is<br />
preferred to superheated water as the heating agent (Wilkinson, 2003). The vessels are<br />
coveredwithcopperdomes thatcarry centrallyplacedflues thatcarry awaysteam.Ineach<br />
decoction some of the mash is pumped to asmaller vessel, (the mash cooker, copper or<br />
kettle) where it is heated to boiling <strong>and</strong> then, after aboil, is transferred back to the mash<br />
mixing vessel (Chapter 4). These vessels can also be used for cooking adjuncts.<br />
Mash cookers are similar in construction to mash-mixing vessels, except that often<br />
stirrers are more powerful <strong>and</strong> the vessels have relatively greater steam heated areas,<br />
since their contents must be heated to boiling. An internally mounted heater may<br />
supplement the heating supplied by a steam jacket. In some small breweries the mash<br />
cooker also acts as the hop-boiling copper (kettle). Mashing vessels have specified<br />
`duties', for example they must be capable of heating the vessel contents linearly at 0.5 or<br />
1 ëC/min. Some decoction mashing breweries, <strong>and</strong> all breweries using double mashing,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
have separate cookers for cooking raw cereal grits. Some may be heated under pressure,<br />
<strong>and</strong> the higher temperatures achieved are adistinct advantage if rice grits are being used.<br />
Early cookers were often horizontally mounted, cylindrical vessels closed by rounded<br />
ends, heated by asteam jacket <strong>and</strong> equipped with ahelical ribbon stirrer supported by<br />
arms extending from acentral rotating shaft (Scott, 1967). Newer cookers are generally<br />
vertically mounted cylinders closed with arounded base <strong>and</strong> cover, heated by asteam<br />
jacket <strong>and</strong>, sometimes, direct steam injection.<br />
6.4.2 Temperature-programmed infusion mashing<br />
Temperature-programmed infusion mashing is less costly than decoction mashing in<br />
terms of number of vessels <strong>and</strong> energy used, decoction mashing needing 20±50% more<br />
energy (steam). Estimates of typical energy use levels are, for programmed infusion<br />
mashes 8.5MJ/hl wort, for single decoction mashing 11.0MJ/hl <strong>and</strong> for double decoction<br />
mashing, 11.6MJ/hl. Mashing in amash mixing vessel (with or without aprogrammed<br />
rise in temperature) <strong>and</strong> separating the wort in alauter tun or mash filter is amore rapid<br />
way of obtaining wort than using amash tun. The greater speed of processing is achieved<br />
because in two-vessel systems asecond mash may be in preparation while the first is<br />
being lautered or filtered. It is not surprising that brewing <strong>practice</strong>s in mainl<strong>and</strong> Europe<br />
<strong>and</strong> the UK are converging on the use of infusion mash mixing vessels <strong>and</strong> lauter tuns or<br />
mash filters. In addition there has been aconvergence in vessel designs, incorporating<br />
refinements that confer increased heating efficiency, ease of use, reduction in shear by<br />
stirring, flexibility, <strong>and</strong> reductions in oxygen uptake.<br />
Modern mash mixing vessels, mash cookers, cereal cookers <strong>and</strong> temperatureprogrammed<br />
mash mixing vessels are very similar. They often have higher height/width<br />
ratiosthanoldervessels<strong>and</strong>commonlyhavetwoorthreeheatingzones,allowingdifferent<br />
volumesofmashtobeheatedefficiently(Barnes<strong>and</strong>Andrews,1998;BuÈhleretal.,1995;<br />
Herrmann, 1998; Kunze, 1996; McFarlane 1993; Wilkinson, 2001, 2003; Wilkinson <strong>and</strong><br />
Andrews, 1996; Figs 6.11, 6.12). Probably in all newer vessels stainless steel is the only<br />
metal in contact with the mash. Steam jackets were all double walled <strong>and</strong> later the walls<br />
were`dimpled'.Nowsteamisoftensuppliedtotheheatingsurfacesinsemi-circularpipes<br />
welded onto the heat exchange surfaces (cf. Chapter 10). The turbulence in the steam in<br />
the pipes improves its heat transfer efficiency <strong>and</strong> so the heating performance <strong>and</strong>, in the<br />
event of sudden cooling or other loss of pressure, the heating system is not so liable to<br />
collapse under vacuum as are double-walled units. Steam heating is applied to the base of a<br />
vessel <strong>and</strong> to the sides. The side <strong>and</strong> base heating is applied in zones that are operated<br />
separately so only the zone(s) covered with mash are heated. This arrangement allows<br />
mashes having different volumes to be processed in one vessel. Sometimes, when direct<br />
steam injection is used for heating, the tangential injection of steam helps to mix the mash.<br />
Burning on a heat exchange surface cannot occur but steam condensation slightly dilutes<br />
the mash, local `overheating', with enzyme destruction, must be a risk <strong>and</strong> the steam must<br />
be `pure' <strong>and</strong> carry no odorous or other contaminants.<br />
Stirring a mash or transferring it between vessels can create shear which damages the<br />
grist, breaking up the particles, extracting more -glucan <strong>and</strong> perhaps altering the structure<br />
of the particles <strong>and</strong> making them more gelatinous. One consequence of shear-induced<br />
damage is that wort separation is slowed. Modern stirrers create less shear than the older<br />
types in which propellers with relatively small diameters turned rapidly. Newer stirrers are<br />
larger, extend across about 85% of the vessel diameter, <strong>and</strong> move more slowly. They cause<br />
less shear <strong>and</strong> create a minimal surface vortex, mix the mash well, <strong>and</strong> sweep the mash<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
Condensate<br />
Condensate<br />
(b)<br />
Pre-masher<br />
Liquor<br />
Condensate<br />
Grist<br />
CIP<br />
Vapour<br />
stack<br />
over the heat exchange surfaces causing local turbulence so optimizing heat exchange.<br />
Scaling <strong>and</strong> local overheating are minimized <strong>and</strong> so there is little consequent enzyme<br />
destruction <strong>and</strong> burning on to the heat exchange surfaces (Figs 6.11, 6.12). In some cases<br />
the need for baffles (used to prevent the mash rotating with the stirrer, but inevitably<br />
causingshear)is overcomebymountingthestirreroff-centre(say5ëfrom thevertical)in<br />
the asymmetrical, dished base of the vessel (Scott, 1967; Wilkinson <strong>and</strong> Andrews, 1996).<br />
Shear during the transfer of mash between vessels in minimized by using gentle<br />
gravity feed where possible or carefully rated, slow-running pumps with wide `throats'<br />
<strong>and</strong> wide piping, avoiding bends where possible using gentle curved bends where<br />
CIP<br />
Outlet<br />
Agitator drive<br />
Inspection<br />
port<br />
Rotation<br />
tip speed<br />
3.8 m/s<br />
maximum<br />
Maximum<br />
level<br />
Steam<br />
Steam<br />
Steam<br />
Fig. 6.11 (a) Asection of modern mash-mixing vessel. Note the inclined base (after Wilkinson<br />
<strong>and</strong> Andrews, 1996). (b) A plan view of the interior of the base of a mash-mixing vessel, showing<br />
the offset stirrer.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 6.12 An alternative pattern of mash-mixing vessel (after Narziss <strong>and</strong> others, 1992).<br />
essential,<strong>and</strong>minimizingthelengthsofpipework.Whileolderstirrerscreatedsubstantial<br />
vortices, which drew air down into the mash, <strong>and</strong> so favoured oxidations, new stirrers do<br />
not, <strong>and</strong> often they run at ahigh speed only during mashing in <strong>and</strong> during heating periods<br />
but at aslower speed at other times. Freshly made mash is often delivered to the base of<br />
the receiving vessel <strong>and</strong> rises in the vessel with little turbulence, rather than being<br />
dropped from above. Both of these refinements reduce oxygen uptake <strong>and</strong> so reduce<br />
oxidationinthemash.Itislesscostlytoincreasethetemperatureofamashbyaddinghot<br />
water rather than by heating with steam (Sommer, 1986). Thus at the lower temperature<br />
thecooler,thickermashfavoursproteolysiswhileinthehotter,thinnermash,obtainedby<br />
the addition of hot water, amylolysis is favoured. It is amusing that this `new' proposal<br />
follows the older British <strong>practice</strong> of underletting with hot water in raked mash tuns. New<br />
mashing vessels are now all equipped with the plumbing needed for automated cleaning<br />
(CIP). CIP has usually been retro-fitted to older equipment. The steam flues of some<br />
mashing vessels, especially mash coppers, may be equipped with economizers to recover<br />
heat, as is the case with hop coppers (Chapter 10).<br />
6.5 Lauter tuns<br />
Lauter tuns have been used for many years to separate worts from decoction mashes.<br />
Now they are also used with temperature-programmed infusion mashing <strong>and</strong> double<br />
mashing systems. In the UK, in a few instances, lauter tuns have been operated as mash<br />
tuns. In recent years the need to accelerate brewing processes, to minimize oxygen uptake<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>and</strong> to improve extract recoveries while obtaining spent grains with lower moisture<br />
contents, <strong>and</strong> to produce equipment to compete with mash filters (Section 6.6) has led to<br />
many refinements in the designs <strong>and</strong> methods of operation of lauter tuns <strong>and</strong> in<br />
improvements in their performances (Andrews <strong>and</strong> Wilkinson, 1996; De Clerck, 1957;<br />
Herrmann et al., 1990; Kunze, 1996; Lenz, 1989; Narziss, 1992; Wilkinson, 1993, 2003).<br />
In general layout lauter tuns resemble mash tuns <strong>and</strong> are usually circular in cross-section<br />
(Fig. 6.13), although rectangular lauter tuns have been used. They are covered with<br />
domes connected to chimneys that carry away steam, <strong>and</strong> they are heavily insulated.<br />
There are significant differences between mash tuns <strong>and</strong> lauter tuns. In particular the<br />
gristsusedinthemashesthatarefilteredinlautersaremorefinelyground,themashesare<br />
more dilute, `thinner', <strong>and</strong> because more rapid wort filtration is desired the bed depth is<br />
comparatively shallow, so it provides aless good filter bed. For agiven capacity, alauter<br />
tun may have a50% greater diameter than amash tun. The bed depths used are very<br />
variable, for example, 35cm (13.8in.), <strong>and</strong> 46cm (18.1in.). Maximum bed loadings may<br />
beinthe range339±153kg/m 2 for cycles allowing6±12 brews/24h(Wilkinson,2003). In<br />
the older system mash was dropped into the lauter tun from above through apipe curved<br />
to minimize the impact on the plates <strong>and</strong> the layer of water covering them. In modern<br />
<strong>practice</strong> the mash is allowed to enter gently through one or more ports either in the sides<br />
of the tun or up through the base, to reduce turbulence <strong>and</strong> the entrainment of air.<br />
Possiblythemashmaybeloadedunderablanketofcarbondioxidegastoexclude air<strong>and</strong><br />
atop-pressure of carbon dioxide can accelerate wort run off (Stippler et al., 1994).<br />
During transfer to the lauter shear should be minimized, for example, by using aslowrunning<br />
pump with an open-throated impeller, giving atransfer velocity of less than<br />
1.3m/s (4.26ft./s). More entry ports enable loading to be carried out faster, so saving<br />
time. The loading that may be used varies with the way in which the grist has been<br />
milled. For dry milled grist the loading on the plates may be 160±175kg/m 2 , for<br />
conditioned <strong>and</strong> milled grist 170±210kg/m 2 ,while for steep-conditioned <strong>and</strong> milled grist<br />
the values are 200±280 or even 310kg/m 2 (Kunze, 1996; Lenz, 1989).<br />
The false bottoms were originally made of brass or gun-metal plates, but now are of<br />
stainless steel or are substituted by stainless steel, profiled wedge wire (Figs 6.8, 6.9).<br />
These plates may be lifted for cleaning, a laborious process that, as with mash tuns, had to<br />
be carried out after every run but now, with modern CIP installations, may need to be<br />
carried out only weekly or even less frequently. While some plates may have drilled<br />
circular perforations, most have slots typically 30±40 mm (about 1.2±1.6 in.) long <strong>and</strong><br />
0.6±0.7 mm (0.024±0.028 in.) wide. As with mash tun plates, these slots widen out below.<br />
There may be 2500±3000 slots/m 2 of plate area. Single milled plates may have a free area<br />
of 6±8%, double milled plates up to 12% <strong>and</strong> wedge wire `plates' of 18% <strong>and</strong> even up to<br />
25%. The plates provide little resistance to the flow of wort, relative to the resistance<br />
provided by the bed of grain, <strong>and</strong> run off rates are the same with plates of 12 or 18% free<br />
area (Lenz, 1989). Run off times may be 1±2 h.<br />
The true bottom of a lauter tun may be flat or consist of a series of depressions or<br />
`valleys' that assist the drainage of wort into the collecting tubes while reducing the<br />
under-deck volume, perhaps by 30%. This reduces the volume of hot water needed to<br />
cover the plates, the amount of washing water required <strong>and</strong> hence the volume of effluent.<br />
The depth of the under-deck space is usually about 20 mm (0.70 in.). The deeper this<br />
space the more water is used in driving out the air, covering the plates <strong>and</strong> heating the<br />
unit, <strong>and</strong> so the more the initial wort is diluted. The wort collection tubes in the true base<br />
may be spaced 1/1.2±1.5 m 2 . The tubes are now usually joined into the base through<br />
conical extensions that reduce the local flow rate across the under-deck space <strong>and</strong> `even<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Insulation<br />
Spent<br />
grains<br />
discharge<br />
Spent grains<br />
receiver<br />
Mash<br />
supply<br />
pipe<br />
Supports<br />
CIP<br />
Sparging<br />
spray<br />
Wort<br />
collection<br />
tank<br />
Vapour<br />
stack<br />
Pillar supporting<br />
rakes<br />
Motor <strong>and</strong> machinery<br />
for rotating, raising <strong>and</strong><br />
lowering the lauter arms<br />
Lighting<br />
Wort pump<br />
Lauter arms with<br />
attached ‘zig-zag’<br />
knives<br />
Supporting pillar<br />
Inspection<br />
port<br />
Run-off pipes<br />
Wort<br />
collection<br />
line<br />
Fig. 6.13 A section of a modern lauter tun equipped with fixed sparging nozzles <strong>and</strong> zig-zag, chevron-shaped knives,<br />
the outer ones of which are the double-shoe type (after Herrmann et al., 1990).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Spent grain<br />
removals<br />
beam,<br />
lowered<br />
position
Perforated false bottom<br />
Under-deck space<br />
True base<br />
Wort collection tube,<br />
with a conical inlet<br />
Fig. 6.14 Adiagram of aconical entry to awort collection tube in the base of alauter tun (after<br />
various sources).<br />
out' the suction through the false bottom <strong>and</strong> on to the bed of grains above (Fig. 6.14). In<br />
the past, as with traditional mash tuns, each pipe carried the wort down to atap with a<br />
`swan neck' that discharged it into acollecting trough or `grant'. The flow of the wort<br />
was regulated by h<strong>and</strong> <strong>and</strong> the clarity <strong>and</strong> gravity was checked. In newer tuns the wort is<br />
usually collected in ring mains or directly into asingle vessel. These arrangements avoid<br />
exposure to the air <strong>and</strong> are necessary for automated operation.<br />
The mashing machinery consists of rotating beams that carry downwardly directed<br />
knives <strong>and</strong>, often, asweep arm or`plough' for removing spent grains. The machinery can<br />
be raised or lowered <strong>and</strong> this is usually automatically controlled. Knives may have<br />
`winglets' which project from the sides. As the knives cut through the mash they lift it<br />
slightlybecause the wingletsare angledtothe horizontal (Fig.6.15a), looseningthemash<br />
<strong>and</strong> assisting run off. The bottom of each knife terminates in a`shoe' which, like the<br />
`winglets' lifts the mash. In some tuns the knives may be rotated through 45±90ë so that,<br />
as the rotating machinery is lowered into the spent grains at the end of amash, these are<br />
pushed into the freshly opened discharge ports. In some lauter tuns the direction of<br />
rotation of the machinery can be reversed <strong>and</strong> this automaticallyalters the attitudes of the<br />
knives from their cutting to their grain discharging angles. Because of their greater<br />
efficiency discharge is increasingly by sweep arms that are lowered into the grain bed in<br />
front of the knives (Fig. 6.16). Greater efficiency of grains removal results in less<br />
pollution of cleaning water <strong>and</strong> so smaller effluent charges (Barnes, 2000).<br />
Several other types of knives are in use. Rather than being straight, some have `zigzag'<br />
blades that cut longer channels in the mash but which, because they are not simple<br />
vertical slots, are less likely to encourage channelling with consequent reductions in<br />
extract recovery. Another development is the `double shoe' knife (Fig. 6.15b). Neither of<br />
thesetypes liftthe mash.Inall cases the knives arespacedalongthe carrier beamssothat<br />
they travel between the tracks of the other knives, the most efficient arrangement for<br />
raking all parts of the grain bed (Fig. 6.17). Knives are arranged so that each cuts the<br />
same area of the mash bed. With large lauter tuns this requires special arrangements of<br />
the arms supporting the knives (Figure 6.17). Inevitably the further knives are positioned<br />
from the centre of the tun the faster they must move through the mash. Sparging liquor<br />
may be supplied from rotating overhead sprays, either free or mounted on the beams of<br />
the raking machinery, or from fixed overhead ducts. The equipment is washed with hot<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(b)<br />
(a)<br />
Upright<br />
Double shoe<br />
Fig. 6.15 A lauter tun knife equipped with `winglets' to lift the mash (after Andrews <strong>and</strong><br />
Wilkinson, 1996). Note the plastic base to the `shoe' which minimizes the risk of damage to the<br />
false bottom of the tun. (b) A pattern of double shoe lauter tun knife with a `zig-zag', chevronshaped<br />
blade (Lenz, 1989).<br />
water at the end of each run, the false bottom being cleaned with high-pressure jets <strong>and</strong><br />
the under-deck space being washed with fixed sprays. CIP with strong cleaning solutions<br />
may be needed only once a week.<br />
A novel lauter tun has an annular filtration area, the central region being occupied with<br />
a cylinder, up through which the shaft carrying the mashing machinery passes (Putman,<br />
2002; Wasmuht et al., 2002). It is argued that extract recovery from the central region of<br />
a conventional tun is inadequate, but since extract recoveries of 99.5% have been<br />
achieved in such tuns, this claim may be doubted. Many details of this tun have been<br />
optimized <strong>and</strong> are claimed to favour the rapid collection of high-gravity worts.<br />
Lauter tuns are increasingly being automated, but whether operated manually or<br />
automatically the stages of an operation cycle are the same. The first operation is to warm<br />
the vessel <strong>and</strong> flood the under-deck space with liquor at the mashing-off <strong>and</strong> sparging<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
Plough bar in rest position<br />
Plough bar in graining-out position<br />
(b)<br />
Grain outlet valves<br />
Following sweep arm<br />
Spent grains sweep arm<br />
Fig. 6.16 (a) A lauter tun with the grain discharge plough bar in the raised (rest) <strong>and</strong> lowered<br />
(graining-out) working positions (after Barnes, 2000). (b) A plan view of another pattern of grain<br />
plough with a trailing, second sweep arm (after Herrmann et al., 1990).<br />
temperature (often 75±78 ëC; 167±172.4 ëF) to drive out the air <strong>and</strong> to create a shallow<br />
layer of water over the deck (1.2±2.5 cm; 0.5±1 in.) to help spread the mash. The mash is<br />
delivered from the mashing vessel, preferably to the sides or up from below, onto the<br />
false bottom. In the past loading was from above. The mash must be uniformly well<br />
mixed. The delivery period is `wasted time' <strong>and</strong> so is shortened as far as possible, for<br />
example, by using several large delivery ports. The freshly loaded mash may be raked a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a) (b)<br />
Fig. 6.17 Possible arrangements of knives on the rotating arms in (a) asmall lauter tun <strong>and</strong> (b) a<br />
large lauter tun. The knives rotate clockwise <strong>and</strong> are so spaced along the bars that they track<br />
between the slots made by the knives that precede them (Andrews <strong>and</strong> Wilkinson, 1996).<br />
few times to help ensure even loading <strong>and</strong> then it is usually (but not invariably) allowed<br />
to st<strong>and</strong> for 5±30min., when some settling occurs. The first layer to form is the shallow<br />
Unterteig, about 1cm (0.39in.) deep, consisting of comparatively large <strong>and</strong> dense grits,<br />
which are often under-modified, starch-rich fragments of malt. Then the bulk of the<br />
grains settles as the Hauptteig, <strong>and</strong> finally athin layer of very fine particles, the Oberteig<br />
isformed. This last layerprovides adisproportionately large part of the resistancetowort<br />
flow <strong>and</strong> which is first raked or `knived' to accelerate wort collection. Above the settled<br />
layer of goods alayer of wort forms. In the past some wort was collected directly from<br />
this upper layer, which could usefully be filtered through a250 min-line filter to<br />
remove floating particles (Berthold, 2000). This process is now rarely used.<br />
The stirred <strong>and</strong> water-logged mash does not float (in contrast to an infusion mash in a<br />
mashtun).Thefirstwortiswithdrawn<strong>and</strong>,becauseitcarriesfinelydividedmaterialsfrom<br />
beneath the false bottom, it is turbid <strong>and</strong> so is re-circulated to the top of the mash where it<br />
is discharged, preferably under the surface of the liquid to reduce the pick-up of oxygen.<br />
The grain bed acts as afilter. When the wort becomes clear, after 5±10min., the flow is<br />
diverted to thecopper (kettle) or acollection vessel. Inmodern <strong>practice</strong>, designed tospeed<br />
up lautering, wort re-circulation may begin when about 50% of the mash has been<br />
transferred. When a filter layer of spent grains is established <strong>and</strong> the wort is clear<br />
collection begins. The first worts are collected until the surface of the grains appears, then<br />
sparging begins. If wort run-off becomes too slow <strong>and</strong>/or the pressure drop across the bed<br />
becomes too high raking will be used, beginning with shallow cuts into the top of the bed,<br />
through the Oberteig, but with progressively deeper cuts as time passes. If the wort<br />
cloudiness increases too much the depth of cutting is reduced. Probably, the continuous<br />
applicationofspargeliquorisusualbutintermittentapplications(twoorthreeapplications<br />
of known volumes) may be more efficient at recovering extract. If the mash sets <strong>and</strong> runoff<br />
ceases an underlet <strong>and</strong>/or sparge may be used, <strong>and</strong> the mash may be `re-slurried' using<br />
the mashing machinery. Sparging may take 2±3h. but is usually less (Table 6.2).<br />
After wort collection, which is stopped when the wort gravity has fallen to some<br />
chosen value, the grains are allowed to drain for about five minutes, to reduce their<br />
moisture content. The discharge ports are then opened <strong>and</strong> the grains are discharged in<br />
ten minutes. or less. Efficient <strong>and</strong> rapid discharge is required, so several large discharge<br />
ports may be provided. The drainings <strong>and</strong> vessel washings contain heat, are rich in<br />
suspended solids <strong>and</strong> have ahigh COD so they are costly to dispose of as effluent.<br />
Furthermore, they contain valuable extract although in dilute solution <strong>and</strong> of alow<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 6.2 An example comparison between a mash tun, a lauter tun <strong>and</strong> a modern mash filter<br />
(Geering, 1996)<br />
Mash Lauter tun 2001 Mash<br />
tun (modern, 1996) filter<br />
Mashing rate (hl/100 kg) 2.8 3.0 2.9<br />
Sparging rate (hl/100 kg) 4.2 3.8 2.4<br />
Total liquor/grist (hl/100 kg) 7.0 6.8 5.3<br />
Filtration area* (m 2 ) 50 90 708<br />
Bed depth* (m) 0.9±1.2 (2 max.) 0.3±0.5 0.03±0.06<br />
Bed loading* (kg/m 2 ) 400 160±220 28<br />
Capacity range (% normal) 60 to +10 50 to +20 20 to +10<br />
Initial wort haze (ëEBC) 10 > 20 > 20<br />
Average wort haze(ëEBC) 4 5±8 < 2<br />
First wort gravity (ëSacch) 80 94 100<br />
Copper gravity (ëSacch) 61 67 70<br />
Lipid content (ppm) 2 18 17<br />
Solids (Imhoff cone, ml/l) < 8 < 12 < 5<br />
Polyphenols (ppm) 165 180 195<br />
Extract recovery (% lab. extract) 96±97 98±99 102<br />
Moisture draff (%) 81 76 64<br />
* For 20 t mashes.<br />
Mash tun cycle: mash in, 20 min.; st<strong>and</strong>, 75 min.; run off, 185±330 min.; drain <strong>and</strong> unload draff, 20 min. Total<br />
time, 300±440 min. Lauter tun cycle: underlet, 3 min.; fill, 11 min.; re-circulate first wort, 4 min.; collect first<br />
wort, 41 min.; second wort, 74 min.; last wort, 10 min.; weak worts, 16 min.; drain, 8 min.; grains out, 25 min.<br />
Total, 192 min. 2001 Mash filter cycle time: filling, 4 min.; filtration, 20 min.; pre-compression, 3 min.;<br />
sparging, 46 min.; compression, 8 min.; drain <strong>and</strong> grains out, 34 min. Total, 115 min.<br />
quality, being relatively poorly fermentable <strong>and</strong> rich in polyphenols <strong>and</strong> lipids. This<br />
liquid, together with lauter tun rinsings, can be stored hot for short periods <strong>and</strong> used in the<br />
mashing liquor for a subsequent mash, sometimes after a treatment with active charcoal<br />
or PVPP. The quantities of solids in the drainings <strong>and</strong> rinsings should be reduced by<br />
ensuring that the grain discharge is thorough, so fewer solids remain to rinse away. It has<br />
been suggested that the liquid should be filtered through a 30 m screen. The possibility<br />
of further purification through a 100 nm pore-size, cross-flow membrane filter with<br />
consequent reductions in the polyphenols <strong>and</strong> the virtual elimination of lipids <strong>and</strong><br />
suspended solids also exists (Barnes, 2000).<br />
The operation of lauter tuns is often fully automated. The flow of wort <strong>and</strong> its<br />
cumulative volume, sparge liquor temperature, volume <strong>and</strong> flow, wort turbidity (haziness),<br />
the pressure differences across the grain bed (between the top of the vessel contents <strong>and</strong> the<br />
under-deck space), <strong>and</strong> between the under-deck space <strong>and</strong> the end of the collection pipe or<br />
the central collection vessel, <strong>and</strong> the positions of the knifing machinery (rotating speed, or<br />
stationary, height, direction of movement, knife setting, setting of grain discharge plough)<br />
are all determined <strong>and</strong> the measurements are fed to a computer. Increases in the cross-bed<br />
pressure instigate deeper cutting into the grain bed while increases in wort turbidity initiate<br />
reductions in the depth of cut. Because lautering is often the slowest process stage in wort<br />
production, there is a constant pressure to reduce wort separation times. The number of<br />
cycles that could be achieved with a lauter tun in 24 hours has risen from six to ten, with 12<br />
being routinely achieved in some instances <strong>and</strong> even 15 cycles/24 h being claimed.<br />
However, for these high rates of use to be achieved low bed loadings must be used.<br />
These rates are dependent on the design of the tun, the nature <strong>and</strong> grind of the grist<br />
being used, as well as the manner in which the mash is carried out <strong>and</strong> the way in which<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the lauter tun is operated. Operating more quickly gives more turbid worts <strong>and</strong> reduces<br />
extract recoveries. The use of inert gases, nitrogen or carbon dioxide, to minimize<br />
oxidative deterioration <strong>and</strong> reduce the oxidation of gel proteins, which contribute to the<br />
flow-resistance provided by the Oberteig, has been combined, at least experimentally,<br />
with the use of gas pressure to accelerate wort run off (Lee et al., 1997; Stippler et al.,<br />
1994, 1995; Stippler <strong>and</strong> Johnstone, 1994). This requires a lauter tun to be fitted with<br />
valves <strong>and</strong> seals to contain the pressure <strong>and</strong> to prevent an excessive pressure increase or<br />
the formation of a vacuum. Pressure-lautering may also give better extract recoveries <strong>and</strong><br />
spent grains containing less moisture.<br />
6.6 The Strainmaster<br />
The Strainmaster, or Nooter tun, was developed as a compact <strong>and</strong> exceptionally rapid<br />
lautering device in the late 1950s (Irvine, 1985; Narziss, 1992; Vermeylen, 1962). Each<br />
unit consisted of a cylindrical or rectangular hopper-bottomed tank, with discharge doors,<br />
`bomb doors', in the base. Each tank had seven layers of stainless steel strainer tubes,<br />
with pear-shaped cross-sections in the lower half of the vessel. The tubes (130 mm<br />
(5.25 in.) high <strong>and</strong> maximal width 50 mm (about 2 in.)) were perforated with vertical<br />
slots, 0.63 mm (0.025 in.) wide <strong>and</strong> 12.7 mm (0.5 in.) long. Each layer of tubes was<br />
connected by a main to a separate wort pump then to separate `swan neck' tubes, which<br />
discharged into a grant (Fig. 6.18). The tun was used with a finely milled grist made into<br />
a thin mash (2.5 hl/100 kg). After preheating the vessel a mash was delivered through two<br />
inlets, which allowed fast <strong>and</strong> even loading. As each layer of tubes was covered by the<br />
mash the appropriate pump was switched on <strong>and</strong> cloudy wort was drawn off <strong>and</strong> recirculated<br />
until the grains formed a filter layer on the slots. As soon as the wort ran bright<br />
it was directed to the copper. When the mash transfer was complete the continued wort<br />
withdrawal caused the volume of the mash to contract. When the volume had declined to<br />
Grant<br />
Vorlauf*<br />
pump<br />
Vorlauf* line<br />
Sewer line Wort to kettle<br />
* Vorlauf = first wort (re-circulation)<br />
Strainmaster<br />
Vapour vent<br />
Mash-transfer<br />
pump<br />
Mash<br />
Sparge water pump<br />
Sparge water<br />
Spent-grain gates<br />
Spent grain to spent-grain receiving tank<br />
Five wort pumps<br />
Five wort draw-off lines<br />
Fig. 6.18 A Stainmaster lauter tun. The patent is owned by Anheuser-Busch (after Briggs et al.,<br />
1981).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
apre-selected value sparging began automatically. When wort collection was complete<br />
cold water might be added to the top of the grains, the discharge doors were opened <strong>and</strong><br />
the spent grains fell <strong>and</strong> were rinsed into acollecting vessel.Various performance figures<br />
were quoted. For example, pre-heating the vessel <strong>and</strong> transferring the mash (6.8t) 30±<br />
40min.; run off, 90min.; grains out, 5min.; flush, 10min., total time 135±145min.<br />
Others claimed turn round times of 105min. at adate (1985) when alauter tun cycle<br />
might be 180min. The performance of this exceptional device, which created an<br />
important benchmark for rapid wort recovery, has now been matched or overtaken.<br />
Problems associated with its use included the production of exceptionally wet grains<br />
(moisture contents over 80%) which necessitated dewatering with presses or vibrating<br />
screens. The press liquor was used to make subsequent mashes. Extract recoveries were<br />
generallylow,rarelyexceeding96%.Wortswerecloudy<strong>and</strong>effluentloadingswerehigh.<br />
The system was not suitable for making high-gravity worts.<br />
6.7 Mash filters<br />
The use of low-pressure mash filter presses was suggested as long ago as 1819, but the<br />
first practical experiments were not carried out until 1874, by Gall<strong>and</strong>, <strong>and</strong> the first<br />
commercialpresswasmadebyMeurain1890(Dixon,1977).Sincethattimemashfilters<br />
have undergone many refinements <strong>and</strong>, since various types are in use, examples will be<br />
described in turn. Filters compete with lauter tuns <strong>and</strong> so comparisons between these<br />
devices will be made. All filters retain the spent grains with filter cloths. Depending on<br />
the cloth used the grist can be moderately, finely or very finely ground, permitting better<br />
extractrecovery.Becausethefilterlayersofgristarethinwortcollectionisrapid(despite<br />
the resistance of the fine grist to liquid flow), <strong>and</strong> so wort collection times are short.<br />
Traditional filters consist of two types of units (plates <strong>and</strong> frames), made of iron or<br />
stainless steel, suspended alternately along aframe <strong>and</strong> pressed together when in use<br />
(Figs 6.19, 6.20. De Clerck, 1957; Dixon, 1977; Irvine, 1985; Kunze, 1996; Narziss,<br />
1992). The plates from which the wort is collected were originally of grooved iron but<br />
these have been replaced by lighter grid plates or folded-sheet plates. The plates have<br />
filter cloths draped over them, which provide the support for <strong>and</strong> retain the mash (Fig.<br />
6.20). In the past cotton cloths were used. These have been replaced by monofilament<br />
polypropylene cloths, which last longer (five times or more) <strong>and</strong> are easier to clean.<br />
Thorough cleaning with caustic solutions may be needed only every 800 brews. In<br />
contrast, cotton cloths needed cleaning after every brew but gave marginally less cloudy<br />
worts. The other units, the frames, are hollow <strong>and</strong> contain the mash. When the stacks of<br />
alternating units, which may be 1.5 m square (4.92 ft. square), are pressed together, often<br />
by a hydraulic ram, the gaskets seal between them so that a series of alternating mashcontaining<br />
<strong>and</strong> wort-collection chambers are formed <strong>and</strong> the holes around the edges of<br />
the units are joined to form channels for the mash, the sparge liquor <strong>and</strong> the wort.<br />
Filters may contain around nine tonnes of grist <strong>and</strong> consist of stacks of 10±60 units.<br />
The bed thickness is chosen with reference to the grist to be used, so the finer the grind<br />
the narrower the bed. Values of 4±6 cm (1.57±2.36 in.) <strong>and</strong> even up to 10 cm (3.94 in.)<br />
have been used. The end plates are different <strong>and</strong> seal off channels or provide connections<br />
to the mash-delivery <strong>and</strong> sparge lines. For mash filters to work well, they need to be<br />
completely filled <strong>and</strong> so they should be used with one invariable volume of mash. To<br />
reduce the capacity of a filter it is necessary to introduce blanking plates, which is<br />
inconvenient <strong>and</strong> laborious. These older units are made of metal <strong>and</strong> the heat losses to the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Wort<br />
Water<br />
Flow of<br />
wort<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Flow of water<br />
I II III<br />
Wort pan<br />
To copper Grains conveyor<br />
Fig. 6.19 A schematic longitudinal section through an older pattern of mash filter (Hind, 1940).
Sparge<br />
Hollow<br />
frame<br />
Cloth-covered<br />
grid<br />
Hollow<br />
frame<br />
Filter cloth<br />
covering grid<br />
(a)<br />
Top sparge<br />
channel<br />
Bottom<br />
sparge channel<br />
Hollow frame<br />
(c)<br />
Mash entry<br />
Channel for<br />
mash entry<br />
Air<br />
channel<br />
Wort<br />
channel<br />
surroundings are high, the filters are uncomfortably hot to work near <strong>and</strong> must be<br />
mounted in a well-ventilated area. Filters occupy a smaller floor area than lauter tuns of<br />
similar capacities. To discharge the spent grains the plates <strong>and</strong> frames are separated<br />
automatically <strong>and</strong> in turn, when the spent grains fall from the frames <strong>and</strong> into a collecting<br />
trough from which they are moved by a screw conveyor. The opening <strong>and</strong> grain<br />
discharging operations are relatively slow. To halve the time taken `double filters' are<br />
used, in which the filters are opened simultaneously from both ends. In operation a filter<br />
is preheated to about 80 ëC (176 ëF) with steam or hot water <strong>and</strong> is checked for leaks. The<br />
filter is emptied <strong>and</strong> well-mixed mash is pumped from the mashing vessel <strong>and</strong> is loaded<br />
into the frames, from above in older designs, but from below <strong>and</strong> with minimum<br />
turbulence in newer types, to minimize oxygen uptake.<br />
Grid<br />
(b)<br />
Wort<br />
Spargings<br />
(d) (e)<br />
Bottom sparge in<br />
Spargings out<br />
Fig. 6.20 The arrangement of the essential components of a traditional mash filter (Briggs et al.,<br />
1981). (a) A vertical section showing how the filter cloths hang over the grids <strong>and</strong> alternate with the<br />
hollow frames, which will contain mash. (b) A face-view of a grid. (c) A face-view of a frame.<br />
(d) A stage when the filter is just full of mash <strong>and</strong> the first worts are escaping (solid lines). The next<br />
stage with sparging into odd numbered grids <strong>and</strong> collecting from even numbered grids (dashed<br />
lines). (e) Last stage in sparging.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Pre-filling of filters with an inert gas to displace oxygen has been considered. The air<br />
or inert gas displaced by the mash is vented. When the frames are full of mash the gas<br />
vents are closed while mash is still being transferred into the frames. The taps or outlets<br />
from the plates are opened <strong>and</strong> the wort, which has escaped through the filter cloths, is recirculated<br />
to the mashing vessel until it is clear <strong>and</strong> then is collected. In older mash filters<br />
the flow of wort from each plate was regulated through individual taps leading to swan<br />
necks, which delivered the wort into a wort collection trough or grant.<br />
In modern units the wort is collected into a common main <strong>and</strong> air is excluded. After<br />
the first wort has been collected the sparge liquor, which may have been rendered<br />
oxygen-free, is admitted into the tops of alternate plates, displacing the wort downwards.<br />
When all the wort has been driven out, the wort outlets from these plates are closed while<br />
those of the remaining plates are opened. Now the sparge liquor passes from the first<br />
plates across through the mash in the frames <strong>and</strong> out from the second plates. When<br />
sparging has been sufficient the last of the sparge liquor is driven through the spent grains<br />
by a top pressure provided by compressed air, which also drives out the last of the<br />
spargings. The wort emerges at about 75 ëC (167 ëF). Extract recovery can be as high as<br />
99.5% of the laboratory extract, the first worts are strong (SG 1085±1090) <strong>and</strong> the final,<br />
average values are 1044±1048. Thus filters are useful for high-gravity brewing. The spent<br />
grains have moisture contents in the range of 60±80%, usually around 70±75%. The turn<br />
around times of these filters is fast enough to allow 8 brews/24h. Sometimes the sequence<br />
of production runs must be interrupted for cleaning, which involves rinsing with hot<br />
water <strong>and</strong> treatments with 1±2% caustic soda containing a wetting agent <strong>and</strong> a chelating<br />
agent, such as EDTA, if polypropylene sheets are used.<br />
A novel development is the recessed chamber-plate filter (Nguyen, 1996). These units<br />
are made of reinforced polypropylene, <strong>and</strong> a filter press consists of only this single type of<br />
unit, together with appropriate end plates, supported on a frame. Each recessed chamber<br />
plate, typically 2.1 m (6.89 ft.) square, consists of a single unit with a recess each side into<br />
which the filter cloth fits <strong>and</strong> is retained by an `O-ring' (Fig. 6.21). Unlike other filters, the<br />
cloth does not project from below the units. When the units are assembled the space<br />
between the filter cloths is the mash chamber while the spaces between the cloths <strong>and</strong> the<br />
transverse partitions of the plates make up the wort collection chambers <strong>and</strong> the sparge<br />
inlets <strong>and</strong> outlets. The plates are opened in the usual way to discharge the spent grains<br />
which separate from the two filter cloths. The use of an air purge at the end of sparging<br />
gives relatively dry spent grains, with < 74% moisture. The worts are concentrated, extract<br />
recoveries equal laboratory yields <strong>and</strong> the cycle time is less that 120 min., permitting 12<br />
cycles/24 h. Like other newer filters made primarily of polypropylene, heat losses are less<br />
Recessed chamber plate<br />
Filter cloth<br />
Fig. 6.21 Two recessed chamber plates, seen in vertical section (after Nguyen, 1996). The space<br />
between the cloths is the mash chamber. The wort is collected in the spaces between the cloths <strong>and</strong><br />
the plate partitions.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ii i ii i ii<br />
(a) (b)<br />
Fig. 6.22 Working diagram of ahigh-pressure mash filter (Waesburghe, van, 1979, 1989).<br />
(a) Vertical section of part of a filter series in which vertically grooved, stainless steel plates (ii);<br />
(b) alternate with the filter pockets (i) which contains the mash. The light arrows indicate the mash<br />
loading position. The spent grains were discharged by unsealing the bases of the pockets which are<br />
sealed by pressure against the groove on the metal plates. The bold arrows indicate the direction of<br />
compression used on the filter assembly.<br />
than those from metal filters, giving more pleasant working conditions <strong>and</strong> allowing the<br />
use of more aggressive cleaning agents without risk of corrosion.<br />
Anotherdevice,introducedin1978<strong>and</strong>whichenjoyedlimitedsuccess,istheHP(highpressure)<br />
filter (Waesberghe, van, 1979, 1991; Irvine, 1985; Nguyen, 1996). This filter<br />
consisted of aset of polyester or polypropylene filter pockets, which could be opened or<br />
closed at the base, alternating with metal plates vertically grooved on one side <strong>and</strong> with a<br />
recessinvolvedwithbagclosure<strong>and</strong>sealingontheother(Fig.6.22).Thebagswerefilled<br />
fromabove<strong>and</strong>,afterashortre-circulationperiodforclarification,thewortwascollected.<br />
Because the grists were finely milled the initial worts were bright. The mash was sparged<br />
withwateraddedtothetopsofthepockets,thenthepocketsweregentlycompressedfrom<br />
7±8cm (2.76±3.15in.) thickness <strong>and</strong> finally squeezed more strongly to 3±4cm (1.18±<br />
1.57in.). The pockets were then opened <strong>and</strong> the comparatively dry grains (50±60%<br />
moisture)droppedout.Extractrecoveriesweregood,thewortswereconcentrated<strong>and</strong>the<br />
turn-round time was rapid, about 70min. These filters did not remain in favour, possibly<br />
because of cleaning problems, turbidity in the later `squeezed' worts, difficulties in<br />
obtaining even sparging <strong>and</strong> the complexity of the equipment (Nguyen, 1996).<br />
The newer generation of filters may be termed membrane compression filters. The<br />
first of these is the Meura 2001 filter (Eyben et al., 1989; Hermia <strong>and</strong> Rahier, 1991a,b,<br />
1992; Jones, 1992; Mieleniewski, 1999). The filter is made of alternating chamber<br />
modules with membranes <strong>and</strong> plates, which are covered with filter cloths. Compression<br />
<strong>and</strong> gaskets seal between the units <strong>and</strong> create the necessary channels for the mash, wort,<br />
spargings, etc. The units are made chiefly of reinforced polypropylene <strong>and</strong> are 2by 1.8m<br />
(6.56 by 5.91ft.). The membrane chamber module consists ofathin, grooved plate which<br />
supports two elastic membranes or `diaphragms' (of polypropylene <strong>and</strong> rubber) on each<br />
side that can be inflated with compressed air. The space between amembrane <strong>and</strong> afilter<br />
cloth, which holds the mash, is about 4cm (1.57in.) wide. The mash is made with afine,<br />
dry hammer-milled or awet disc-milled grist. A60-plate filter can accommodate a10.5t<br />
mash, divided into 120 beds <strong>and</strong> resting against 60 double filtration cloths. The stages of<br />
operation are as follows (Fig. 6.23). The filter is pre-warmed <strong>and</strong> then mash is pumped<br />
into the chambers from below, while the air is vented. This upward delivery minimizes<br />
oxygen uptake. When the chambers are full the vents are closed, the wort outlets are<br />
opened, wort collection begins <strong>and</strong> mash delivery is continued until all the mash is in the<br />
chambers. During this process afilter-layer of solids quickly builds up on the filter cloth<br />
<strong>and</strong> the wort becomes bright so quickly that re-circulation is rarely needed. The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
ii
membranes are then inflated with compressed air, (0.5±0.6 bar; 7±9psi), <strong>and</strong> these<br />
squeeze the mash, displacing the solids across the chamber <strong>and</strong> compressing them so that<br />
they adhere to the filter cloths <strong>and</strong> strong wort is squeezed out. Squeezing improves the<br />
homogeneity of the bed, as well as making it thinner, <strong>and</strong> so improves the efficiency of<br />
sparging.Beforesqueezingabout50%oftheextractisretainedinthegrist.Afterthisfirst<br />
compression the membranes are relaxed <strong>and</strong> the spaces, which appear between them <strong>and</strong><br />
the grist layers, are filled with sparge liquor (degassed for preference), usually at about<br />
78ëC (172.4ëF).<br />
When sparging is complete the liquor inlets are closed <strong>and</strong> the mash is compressed a<br />
second time, by inflating the membranes with compressed air (0.7bar; 10psi), <strong>and</strong> the<br />
last spargings are squeezed out. The plates <strong>and</strong> frames are automatically separated in turn<br />
<strong>and</strong> the spent grains fall into areceiving trough with aconveyor. Higher air pressures in<br />
the last compression give grains with smaller moisture contents, but extremely low<br />
moisture contents making the grains difficult to h<strong>and</strong>le <strong>and</strong> values of 60±70% are<br />
probably usual. These filters have rapid turn-round times, allowing 12 cycles/24h, the<br />
worts are strong <strong>and</strong> extract recovery is exceptional, often 101 <strong>and</strong> even 103% of the<br />
laboratory value (Table 6.2). The fatty acid contents of these worts are very low, 10±<br />
24mg/l for Meura 2001 worts, 27mg/l for lauter tun worts <strong>and</strong> even 143mg/l for filter<br />
press worts. It has been claimed that the loading can be reduced to 70% of normal, but<br />
practical difficulties have been noted even at loadings reduced to 85% of the st<strong>and</strong>ard<br />
value.Thisdiscrepancymaybeduetodifferencesinthegristsused.Thisfiltercopeswell<br />
with awide variety of grists, including those containing mostly unmalted sorghum or<br />
roasted <strong>and</strong> flaked barley. The first <strong>and</strong> total worts are unusually concentrated, e.g.,<br />
14.5ëP (SG 1059) <strong>and</strong> 11.8ëP (SG 1048) respectively, <strong>and</strong> so are suited to high-gravity<br />
brewing. Filters are cleaned regularly by rinsing <strong>and</strong> more completely, say weekly, using<br />
caustic cleaning agents <strong>and</strong> possibly hydrogen peroxide. As the filter cloths become dirty<br />
the resistance to wort flow increases. As with other polypropylene filters the heat loss is<br />
much less than with the older, metal filter presses.<br />
The other membrane filter to be described is the MK 15/20, intended for 5±12.5t<br />
mashes (Karstens, 1996; Kunze, 1996; Michel, 1993). Smaller versions have other<br />
reference numbers. The filters consist of alternating membrane <strong>and</strong> chamber plates,<br />
mostly made of polypropylene. Filter cloths are firmly fastened over both sides of both<br />
kinds of plates (Fig. 6.24). When the units are assembled <strong>and</strong> pressed together the edges<br />
of the cloths seal (replacing gaskets) <strong>and</strong> create mash chambers bounded on both sides by<br />
filter cloths, so wort can escape from both sides of the mash. The sequence of operations<br />
is indicated in Fig. 6.24. The mash cycle time is about 120min.<br />
6.8 The choice of mashing <strong>and</strong> wort separation systems<br />
The choice of mashing system depends on many factors. For the production of traditional<br />
beers it may be imperative to continue using near-isothermal mashing in a mash tun or<br />
decoction mashing to retain the character of a product. In the same way, a particular<br />
fermentation system may have to be retained. For small traditional breweries making ales<br />
a mash tun, with its simple construction <strong>and</strong> low maintenance requirements, serves well if<br />
not more that two to three brews ervery 24 hours are required. For larger breweries,<br />
making large volumes of single beers, the economic advantages of using all the brewing<br />
equipment for as much of the week as possible, <strong>and</strong> making as many brews every 24<br />
hours as possible are very great. Because both conversion <strong>and</strong> wort separation occur in<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
(c)<br />
Air<br />
Mash<br />
1st wort<br />
1st wort<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
(d)<br />
(b)<br />
Mash<br />
1st wort<br />
Sparge liquors<br />
Spargings
(e)<br />
the single vessel in amash tun <strong>and</strong> the depth of the bed of grist is large, the turn round<br />
time is slow. Some decoction mashing schedules are slow (Chapter 4), but with<br />
temperature-programmed infusion mashing with an all-malt grist conversion should be<br />
completeintwohours,orless,<strong>and</strong>wortseparationmaybecompletedsufficientlyquickly<br />
to readily allow 8±10 (<strong>and</strong> possibly 12) brews/24hwith alauter tun <strong>and</strong> 10±12 brews/<br />
24h. with anew mash filter.<br />
Both lauter tuns <strong>and</strong> mash filters can produce high-quality worts (Table 6.2). Amash<br />
filter using hammer milled grist will give abetter extract recovery, amore concentrated<br />
wort, drier spent grains, use less water <strong>and</strong> generate less effluent, but is likely to be more<br />
costly than alauter tun of equivalent capacity. If the brewery is making avariety of beers<br />
that necessitate the use of different brew lengths then the greater flexibility of the lauter<br />
tun becomes decisively important. Furthermore, the use of acomparatively coarsely<br />
ground grist, suited to alauter tun, may be needed to retain the character of aparticular<br />
beer. Lauter tuns are likely to be less expensive than newer types of mash filters <strong>and</strong><br />
require less maintenance <strong>and</strong> fewer replacements. The operations of both are now usually<br />
automated. Flexibility may be increased by having mash conversion vessels of different<br />
Air<br />
Sparging<br />
Fig. 6.23 Stages in the operation of aMeura 2001 membrane mash filter (various sources). (a)<br />
The mash being loaded into the frames from below the membrane <strong>and</strong> the filter cloth. Air is being<br />
expelled at the top. (b) All the air has been expelled, the vents have been closed, mash is still being<br />
pumped into the filter <strong>and</strong> the first wort is escaping through the filter cloth <strong>and</strong> is being collected.<br />
(c) All the mash has been transferred into the filter, the inlet lines are closed <strong>and</strong> compressed air has<br />
been directed behind the membranes, pushing them out <strong>and</strong> causing them to squeeze the mash,<br />
releasing entrained first wort. (d) Sparge liquor has been admitted between the deflated membrane<br />
<strong>and</strong> the compressed mash <strong>and</strong> the first spargings are being collected. (e) The membranes have been<br />
inflated for the second time, to squeeze the spargings from the draff. In the next stage the filter is<br />
opened <strong>and</strong> the spent grains fall into a collection trough <strong>and</strong> are conveyed away.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
A<br />
Sp<br />
A<br />
Wort 1 Wort 1<br />
(a) (b) (c)<br />
L<br />
capacities <strong>and</strong> the choice of using either one of two wort separation units (Mieleniewski,<br />
1999). Alternatively, when slow mash conversions occur, for example because relatively<br />
high concentrations of slow-converting mash adjuncts are used, two mash conversion<br />
vessels may alternately deliver mash to a single wort separation unit. It is apparent that<br />
for optimal working all the components used in the preparation of wort, raw materials<br />
h<strong>and</strong>ling, milling <strong>and</strong> grist h<strong>and</strong>ling, mashing/conversion, sweet wort separation, hopboiling,<br />
trub separation, wort cooling <strong>and</strong> fermentation must have the appropriate<br />
capacities <strong>and</strong> working rates to avoid `bottlenecks' <strong>and</strong> permit fast working at periods of<br />
maximum dem<strong>and</strong>.<br />
6.9 Other methods of wort separation <strong>and</strong> mashing<br />
Sp<br />
Various unusual methods of wort separation have been tested (Briggs et al., 1981; Lotz et<br />
al., 1997; Dixon, 1977; BeÂndek et al., 1991; Darling, 1968; Schneider et al., 2001;<br />
L<br />
Sparge<br />
(d) (e) (f)<br />
Fig. 6.24 The operational stages of an MK 15/20 filter (various sources). (a) Mash is added at the<br />
base <strong>and</strong> rises between the two filter cloths, <strong>and</strong> air is displaced, A. (b) All the air has been<br />
displaced, the vents are closed <strong>and</strong> mash entry continues. Layers of grain build up on the filter<br />
cloths <strong>and</strong> escaping wort is collected. (c) Compressed air, A, inflates the membranes <strong>and</strong><br />
compresses the mash. The displaced wort is collected. (d) Sparge liquor, L, is admitted <strong>and</strong> the<br />
spargings, Sp, are collected. (e) Possible sparging between the two filter cloths, but in the reverse<br />
direction. (f) At the end of sparging the grist is squeezed again <strong>and</strong> the displaced spargings are<br />
collected. At the end of this sequence the filter is opened <strong>and</strong> the draff falls into a collecting trough.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
A<br />
A
SchoÈffel <strong>and</strong> Deublein, 1980). With the proposed dynamic disc mash filtration technique<br />
the rotating disc drives the mash tangentially to the filtration surface <strong>and</strong> the wort is<br />
membrane filtered. In the Pablo system the wort is separated from the mash with two,<br />
doublestagedecantingcentrifuges,intheReitersystemarotaryvacuumfilterisused<strong>and</strong><br />
gives a high extract recovery. Other devices tried include vibrating screen filters,<br />
vibrating membrane filters, horizontal <strong>and</strong> inclined belt filters, with or without suction,<br />
cross-flow filters, Archimedean screws working in slotted casings <strong>and</strong> cyclones. Many of<br />
these methods were intended to be used with fractionated malt grists lacking husk<br />
materials so that relatively little draff would remain after the conversion process. Some<br />
were used for short periods but many did not pass beyond the experimental stage.<br />
During the period 1955 to 1975 there was intense interest in continuous brewing, with<br />
its advantages of relatively compact, continuously operating equipment with its steady<br />
dem<strong>and</strong> for services <strong>and</strong> its predicted uniform product quality. Large production plants<br />
were constructed <strong>and</strong> used but many problems were encountered that led to them being<br />
replacedbybatchproductionunits. Atthepresenttimecontinuousfermentationiscarried<br />
out on an industrial scale by only one company in New Zeal<strong>and</strong> (Chapter 14) <strong>and</strong> this is<br />
not linked to continuous wort production. However, there is a revival of interest in<br />
continuous fermentation <strong>and</strong> conditioning using immobilized yeasts <strong>and</strong> this may lead<br />
back to an interest in continuous mashing, since processing is most advantageous when<br />
all stages are continuous (Briggs et al., 1981; Darling, 1968). Perhaps the most promising<br />
continuous mashing system was the rotary table filter proposed by APV (Fig. 6.25). In<br />
this equipment the mash travelled, in sequence in plug flow, through stainless steel tubes<br />
Hot<br />
water<br />
Mash<br />
mixer<br />
Steam<br />
Hot water<br />
Flake Roast Malt<br />
Flake weigh<br />
feeder<br />
Mash<br />
filter<br />
Empty<br />
wash<br />
Prefill<br />
fill<br />
Malt weigh<br />
feeder<br />
Mill<br />
Converter tank<br />
Recycle<br />
Recycle<br />
Recycle<br />
Re-circulation<br />
Re-circulation<br />
Hot water<br />
returned<br />
Wort out<br />
Hot water in<br />
Condensate<br />
Fig. 6.25 A diagram of the A. P. V. continuous mashing <strong>and</strong> lautering system (Briggs et al., 1981).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
held in `converter tank(s)' at the chosen mashing temperature(s). The converted mash<br />
was loaded into one of eight mash buckets, each of which resembled asmall mash tun.<br />
For two hours atun moved around acentral support, occupying each of 16 positions in<br />
turn, for15 min. The positions were used for filling, wort recycling, collection <strong>and</strong><br />
sparging, spent grain discharge <strong>and</strong> cleaning, then return to the filling position.<br />
6.10 Spent grains<br />
The undesirable characteristics of mash <strong>and</strong> spent grain drainings as effluents, with their<br />
high COD, BOD <strong>and</strong> suspended solids values, have been noted (Chapter 3), as has the<br />
nutritional valueofdraffinanimal feeds(Chapter4).Itisbestforthe brewer<strong>and</strong>theuser<br />
that the draff is reasonably dry <strong>and</strong> certainly not seeping liquid. Drier draff reflects better<br />
extract recovery for the brewer <strong>and</strong> smaller transport costs <strong>and</strong> storage problems for the<br />
user. The wet state of the draff from Strainmasters was a major factor in replacing them<br />
with other equipment. Conversely, the low moisture content of the draff from mash filters<br />
is a factor in their favour. The production of dry draff becomes even more important<br />
where local conditions dictate that it must be thoroughly dried. Drying is an expensive<br />
process. Spent grains are usually moved by helical screw or compressed air conveyors to<br />
storage bins which typically discharge their contents directly into lorries. This area of the<br />
brewery must be regularly cleaned, since many microbes multiply on wet grains <strong>and</strong><br />
drainings, so these represent potential sources of contamination <strong>and</strong> product spoilage.<br />
6.11 Theory of wort separation<br />
The theories applied to wort separation have been considered <strong>and</strong> have been of value in<br />
designing new equipment (lauter tuns <strong>and</strong> mash filters) but they apply to `ideal' situations<br />
while, in <strong>practice</strong>, the conditions are often far from perfect (Briggs et al., 1981; Dixon,<br />
1977; Harris, 1968, 1971; Hermia <strong>and</strong> Rahie, 1992; Royston, 1966; Wilkinson, 2001).<br />
The modified Darcy's equation (originally derived from the passage of water through<br />
beds of s<strong>and</strong>) may be written:<br />
V = K A P/L<br />
where V is the rate of liquid flow through the bed of particles, A is the area of the bed,<br />
P is the pressure difference across the bed, K is the permeability of the bed, L is the<br />
length of the path through the bed, or the bed depth, <strong>and</strong> is the viscosity of the liquid. K,<br />
the mash bed permeability, ˆ 3 de 2 /180(1 ) 2 , where ˆ bed porosity (wort volume/<br />
mash volume) <strong>and</strong> de ˆ effective particle diameter (sum of the weight fraction/particle<br />
diameter). In a mash the viscosity will depend on the concentration <strong>and</strong> nature of the wort<br />
<strong>and</strong> the temperature. The use of adjuncts that give viscous worts creates run-off problems.<br />
The porosity of the bed is more important than that of the support, the false bottom of a<br />
mash or lauter tun or the clean filter cloth of a mash filter, <strong>and</strong> is proportional to the mean<br />
diameter squared of the grist particles. Thus the finer the grind of the grist, the smaller the<br />
particles <strong>and</strong> the greater the resistance to flow.<br />
The particle shape is also important. The faster flow of lauter tuns relative to mash<br />
tuns is achieved (despite the finer grists used) by reducing L, the bed depth, <strong>and</strong> this<br />
process is carried further in mash filters. While increasing the pressure can increase the<br />
rate of wort flow this approach can be counter-productive since the mash beds are far<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
from ideal <strong>and</strong> can be compressed to such an extent that flow is reduced or even ceases<br />
<strong>and</strong> a`set mash' is created. As wort separation proceeds so the bed of grains builds up to<br />
amaximum thickness, L, <strong>and</strong> may then decrease as the bed contracts or is compressed.<br />
After the first wort is collected <strong>and</strong> sparging begins wort concentration <strong>and</strong> hence its<br />
viscosity, falls. Similar conclusions can be drawn from modifications of Poiseuille's<br />
equation, derived to describe the flow of liquids through bundles of equal sized<br />
capillaries, in which the diameter of the capillaries is replaced by the mean diameter of<br />
the pores in the bed (or the related value of `voidage'). The flow through the bed is<br />
proportional to the square of the mean diameter of the pores.<br />
As sparge liquor moves into the bed of grist it displaces the wort from between the<br />
particles <strong>and</strong> leaches extract from within them. Efficient leaching is essential if high<br />
yields of extract are to be obtained. The solid phase mass transfer coefficient, K/D/d<br />
whereDisthediffusioncoefficient<strong>and</strong>disthediameteroftheparticle.Thus,thesmaller<br />
the particle the greater the surface/volume ratio <strong>and</strong> the shorter the distances from points<br />
within the particle to the surface <strong>and</strong> the faster extract can be leached from it. Diffusion<br />
occurs faster at higher temperatures but sparging temperatures are limited to about 78ëC<br />
(172.4ëF)byotherconsiderations.Leachingtakestime<strong>and</strong>spargingtoorapidlyresultsin<br />
inadequate leaching <strong>and</strong> areduced recovery of extract. Since leaching is favoured by a<br />
finely ground grist while flow rate is favoured by larger particles, in <strong>practice</strong>, a<br />
compromise grist particle size must be sought for each type of wort separation<br />
equipment. The more brews/day that are required the lower the loading on alauter tun<br />
mustbe.Forexample,fortotalcycletimesof240,180<strong>and</strong>120min.alautertun,usingall<br />
malt grists, had loadings (kg/m 2 )of 339, 246 <strong>and</strong> 153 respectively (Wilkinson, 2001).<br />
During the movement of sparge water through the mash bed there is aprogressive<br />
leaching of soluble substances from within the particles <strong>and</strong> a gradient of wort<br />
concentration is established, which increases downwards through the bed. The theoretical<br />
stages are, firstly, the diffusion of extract from the interiors to the surfaces of the grist<br />
particles <strong>and</strong>, secondly, the movement of the extract into the liquid between the particles<br />
<strong>and</strong> its removal with this flowing liquid. Leaching has been approximately described by<br />
equations based on the number of theoretical washing stages involved (Table 6.3). More<br />
washing stages are required to minimize losses in the preparation of strong worts. Greater<br />
retention of liquor in the spent grains leads to greater losses of extract <strong>and</strong> the need for<br />
more extensive washing. The `squeezes' applied in operating membrane mash filters<br />
reduce, firstly, the amount of strong wort <strong>and</strong>, secondly, the amount of spargings in the<br />
draff <strong>and</strong> so enhance extract recovery. As grain beds become more free running so the<br />
washing efficiency decreases <strong>and</strong> there is a tendency for extract recovery to decline. It<br />
follows that increasing the rate of sparging carries with it the risk of a significant fall in<br />
extract recovery. As previously noted with some plant, like the Strainmaster, the losses of<br />
extract in wet spent grains can be so high that it is desirable to recover the extract from<br />
the grain pressings, with the extra cost <strong>and</strong> effort involved <strong>and</strong> the risk of reducing<br />
product quality.<br />
Another process that occurs during wort separation is the filtration of the fine particles<br />
from the wort. It is desirable that the wort be as bright as possible. The nature <strong>and</strong><br />
abundance of these particles depend on many factors. The large bed depths used in mash<br />
tuns ensure that, combined with wort re-circulation, very clear worts can be obtained.<br />
Generally lauter tuns give less clear worts, in part because of the need to `knife' or `rake'<br />
the bed to maintain an acceptable rate of run off. Older, `classical' mash filters gave rise<br />
to even more hazy worts but the newer, membrane compression mash filters give rise to<br />
exceptionally clear worts. This sequence seems to be the result of a several conflicting<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 6.3 Theoretical losses in yield (%) in the extraction of a thick mash made with a liquor/grist ratio of 2 : 1 (wt/wt; Royston, 1996)<br />
For a wort of SG 1050 (12.4 ëP) For a wort of SG 1100 (23.7 ëP)<br />
Spent grains<br />
moisture First<br />
Washing stages<br />
First<br />
Washing stages<br />
contents (%) worts 1 2 3 4 5 worts 1 2 3 4 5<br />
87 22 6 1.5 0.5 0.1 0.1 40 18 11 8 6 4<br />
75 12 2 0.4 0.1 0.1 0.1 25 11 6 4 3 2<br />
60 7 0.7 0.1 0.1 0.1 0.1 15 5 3 1.5 0.8 0.5<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Extraction (%)<br />
100<br />
80<br />
60<br />
40<br />
20<br />
traits, that thicker beds make better filters, more fines are formed with finer grinding <strong>and</strong><br />
exceptionally finely ground (hammer milled) grists are able to retain very fine suspended<br />
materials, presumably because the mean pore sizes through the beds are very small. The<br />
haziness of worts is also influenced by the grist composition <strong>and</strong> is probably increased by<br />
attempts to collect worts too fast, at rates that exceed the optima of the different pieces of<br />
equipment.<br />
Figure 6.26 illustrates the progressions of recovery of extract during wort run off from<br />
several types of separation devices. The rates are Strainmaster > lauter tun > mash tuns.<br />
Thus, as predicted, the larger the filtration area /unit mash the faster extract collection.<br />
Wort recovery from the mash tun with medium depth <strong>and</strong> an all-malt grist is complete<br />
(97% extract recovery) in 255 min., while the values with the deeper tuns, with malt <strong>and</strong><br />
malt <strong>and</strong> maize grits, (98% extract recovery) are 330 <strong>and</strong> 400 min. The performance (true<br />
filtration efficiency) of the Stainmaster <strong>and</strong> the lauter tuns was not as great as predicted<br />
on theoretical grounds, possibly because, at least in part, the mash beds were compressed<br />
(Harris, 1971).<br />
6.12 References<br />
1 2 3 4 5 6<br />
0 0 1 2 3 4 5 6 7 8<br />
Run-off time (h)<br />
Fig. 6.26 The patterns of the cumulative extract recovery against wort filtration time for various<br />
systems (Harris, 1968). (1) A Strainmaster lautering vessel. (2) A lauter tun loaded with a decoction<br />
mash. (3) A lauter tun loaded with a transferred infusion mash. (4) A mash tun with a medium depth<br />
of mash (1.52 m; 5 ft.). (5) A mash tun with a deep mash bed (2.44 m; 8 ft.). (6) A mash tun with a<br />
deep bed of grist containing both malt <strong>and</strong> maize grits, which slow wort separation.<br />
ANDREWS, J. M. H. <strong>and</strong> WILKINSON, N. R. (1996) Ferment, 9 (5), 257.<br />
BARNES, Z. C. (2000) Ferment, 13 (1), 27.<br />
BARNES Z. C. <strong>and</strong> ANDREWS, J. M. H. (1998) Proc. 5th Aviemore Conf., Malting, <strong>Brewing</strong> <strong>and</strong> Distilling<br />
(Campbell, I. ed.). London, The Institute of <strong>Brewing</strong>, p. 127.<br />
BEÂ NDEK, G., HORVAÂ TH, I. <strong>and</strong> ULLMANN, P. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 633.<br />
BERTHOLD, U. (2000) Proc. 26th Conv. Inst. of <strong>Brewing</strong> (Asia Pacific Section), Singapore, p. 52.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting <strong>and</strong> <strong>Brewing</strong> Science (2nd edn),<br />
Vol. I Malt <strong>and</strong> Sweet Wort. London, Chapman <strong>and</strong> Hall, 387 pp.<br />
BROWN, H. T. (1910) J. Inst. <strong>Brewing</strong>, 16, 112.<br />
BUÈ HLER, T. M., MATZNER, G. <strong>and</strong> MCKECKNIE, M. T. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels,<br />
p. 293.<br />
DARLING, R. O. (1968) Brewers' Guard., Apr., p. 147.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
DE CLERCK, J. (1957) A Textbook of <strong>Brewing</strong>, vol. 1. (Trans. Barton-Wright, K.). London, Chapman <strong>and</strong><br />
Hall, 587 pp.<br />
DIXON, I. (1977) Brewers' Guard., 106 (10), 43.<br />
EYBEN, D., HERMIA, J., MEURENS, J., RAHIER, G. <strong>and</strong> TIGEL, R. (1989). Proc. 22nd Congr. Eur. Brew. Conv.,<br />
Zurich, p. 275.<br />
GEERING, P. (1996) Brewers' Guard., 125 (10), 32.<br />
HARRIS, J. O. (1968) J. Inst. <strong>Brewing</strong>, 74, 500.<br />
HARRIS, J. O. (1971) in Modern <strong>Brewing</strong> Technology (Findlay, W. P. K. ed.). London, MacMillan Press, p. 1.<br />
HERRMANN, H. (1998) Ferment, 11 (1), 36.<br />
HERRMANN, H., KANTELBERG, B. <strong>and</strong> LENZ, B. (1990) Brauwelt Internat., (1), 22.<br />
HERMIA, J. <strong>and</strong> RAHIER, G. (1991a) Louvain Brew. Lett., 4 (3/4), 30.<br />
HERMIA, J. <strong>and</strong> RAHIER, G. (1991b) Louvain Brew. Lett., 4 (2), 24.<br />
HERMIA J. <strong>and</strong> RAHIER, G. (1992) Ferment, 5 (4), 280.<br />
HIND, H. L. (1940) <strong>Brewing</strong> Science <strong>and</strong> Practice. Vol. II. <strong>Brewing</strong> Processes. London, Chapman <strong>and</strong> Hall,<br />
pp. 507±1020.<br />
HOPKINS, R. H. <strong>and</strong> CARTER, W. A. (1933) J. Inst. <strong>Brewing</strong>, 39, 59.<br />
IRVINE, J. A. (1985) The Brewer, Feb., p. 46.<br />
JONES, I. R. (1992) Brewers' Guard., 121 (10), 21.<br />
KARSTENS, W. (1996) Brauwelt Internat., 14 (4), 340.<br />
KUNZE, W. (1996) Technology <strong>Brewing</strong> <strong>and</strong> Malting. (Wainwright, T., transl.). Berlin, VLB, pp. 726.<br />
LEE, B. W., KIM, I. K. <strong>and</strong> SHIN, J. Y. (1997) MBAA Tech. Quart., 34 (2), 75.<br />
LENZ, B. (1989) Proc. 2nd Sci. Tech. Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.),<br />
Johannesburg, p. 250.<br />
LOTZ, M., SCHNEIDER, J., WEISSER, H., KROTTENTHALER, M. <strong>and</strong> BACK, W. (1997) Proc. 26th Congr. Eur.<br />
Brew. Conv., Maastricht, p. 299.<br />
MCFARLANE, I. K. (1993) Ferment, 6 (3), 177.<br />
MICHEL, R. (1993) Brauwelt Internat., (2), 123.<br />
MIELENIEWSKI, A. (1999) Brew. Distill. Internat., 30 (3), 12.<br />
NARZISS, L. (1992) Die Bierbrauerei, Bd. II. Die Technologie der WuÈrzebereitung, (7th edn). Stuttgart,<br />
Ferdin<strong>and</strong> Enke, 402 pp.<br />
NGUYEN, M. T. (1996) Ferment, 9 (6), 329.<br />
PUTMAN, R. (2002) Brewer Internat., 2 (10), 4.<br />
REHBERGER, A. J. <strong>and</strong> LUTHER, G. E. (1994) in H<strong>and</strong>book of <strong>Brewing</strong>, (Hardwick, W. A. ed.). New York,<br />
Marcel Dekker, p. 247.<br />
ROYSTON, M. G. (1966) J. Inst. <strong>Brewing</strong>, 72, 351.<br />
SAMBROOK, P. (1996) Country House <strong>Brewing</strong> in Engl<strong>and</strong>, 1500±1900. London, The Hambledon Press,<br />
p. 311.<br />
SCHNEIDER, J., KROTTENTHALER, M., BACK, W. <strong>and</strong> WEISSER, H. (2001) Proc. 28th Congr. Eur. Brew.<br />
Conv., Brussels, CD, paper 22.<br />
SCHOÈ FFEL, F. <strong>and</strong> DEUBLEIN, D. (1980) Brauwissenschaft, 33 (10, 11), 263, 304.<br />
SCOTT, P. MCM. (1967) Brewer's Guild J., p. 339.<br />
SOMMER, G. (1986) Brauwelt Internat., (1), 23.<br />
STIPPLER, K. <strong>and</strong> JOHNSTONE, J. (1994) Brew. Distill. Internat., Apr., 25, 26.<br />
STIPPLER, K., WASMUHT, K., PRITSCHER, R. <strong>and</strong> KEIM, N. (1994) Proc. 23rd Conv. Inst. of <strong>Brewing</strong> (Asia<br />
Pacific Sect.), Sydney, p. 56.<br />
STIPPLER, K., WASMUHT, H., PRITSCHER, R. <strong>and</strong> KEIM, N. (1995) MBAA Tech Quart., 32 (1), 1.<br />
SYKES, W. J. <strong>and</strong> LING, A. R. (1907) The Principles <strong>and</strong> Practice of <strong>Brewing</strong> (3rd edn) London, Charles<br />
Griffin <strong>and</strong> Co., 588 pp.<br />
VERMEYLEN, J. (1962) TraiteÂe de la Fabrication du Malt et de la BieÁre. 2. G<strong>and</strong>. Assoc. Royale des<br />
Anciens EleÁves del'Institut SupeÂrieur des Fermentations, pp. 751±1624.<br />
WAESBERGHE, J. W. M. VAN (1979) Brew. Distill. Internat., 9 (9), 54.<br />
WAESBERGHE, J. W. M. VAN (1989) MBAA Tech. Quart., 17 (2), 66.<br />
WAESBERGHE, I. R. J. VAN (1991) Proc. 3rd Sci. Tech. Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern<br />
African Section), Victoria Falls, p. 96.<br />
WASMUHT, K., STIPPLER, K. <strong>and</strong> WEINZIERL, M. (2002) Brauwelt Internat., 20 (5), 286.<br />
WILKINSON, N. R. (1993) Brew. Distill. Internat., May, 24, 12.<br />
WILKINSON, N. R. (2001) Brewers' Guard., 130 (4), 29; (5), 22.<br />
WILKINSON, N. R. (2003) Brewers' Guard., 132 (2), 24; (3), 20.<br />
WILKINSON, N. R. <strong>and</strong> ANDREWS, J. M. K. (1996) Ferment, 9 (4), 215.<br />
YAMAGUCHI, I., UEDA, T., UJIHARA, S., YAMADA, M. <strong>and</strong> FUKUSHIMA, S. (1997) Proc. 26th Congr. Eur.<br />
Brew. Conv., Maastricht, p. 257.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
7<br />
Hops<br />
7.1 Introduction<br />
Medieval ale (unhopped beer) rapidly went sour <strong>and</strong> turned into malt vinegar. Many<br />
herbs were used in attempts to prolong the shelf-life of such ale (Johnstone, 1997; Behre,<br />
1999) but only the hop, Humulus lupulus L., is used in large-scale brewing today<br />
although some microbreweries use other herbs. Detailed information about hops is found<br />
in abook by Neve (1991) <strong>and</strong> an earlier book by Burgess (1964). Abooklet, Hops <strong>and</strong><br />
Hop Picking, by Filmer (1982) gives many illustrations. The European Brewery<br />
Convention has published aManualofGood Practice±Hops <strong>and</strong>Hop Products (Benitez<br />
et al., 1997) <strong>and</strong> the Proceedings of two symposia on Hops (European Brewery<br />
Convention,1987<strong>and</strong>1994).Moir(2000)providedamillenniumreview.Hopsaregrown<br />
throughout the world, as illustrated in The Hop Atlas (Barth et al., 1994), solely to meet<br />
the requirements of the brewing industry (Table 7.1); the amounts used by herbalists <strong>and</strong><br />
for hop pillows are negligible. Hops of commerce are the dried cones of the female plant<br />
but today much of the crop is processed into pellets <strong>and</strong> extracts. In Europe the yield of<br />
hops is usually expressed in zenters (1 zenterˆ50kgˆ110.23lb.) but in the USA, the<br />
yield is usually expressed in pounds.<br />
Although hops were probably used first for their preservative value, they introduced<br />
bitterness <strong>and</strong> apleasant flavour, which was liked, <strong>and</strong> which is the reason for their<br />
continued use. These flavours were found to originate mainly in the resins <strong>and</strong> essential<br />
oils found in the lupulin gl<strong>and</strong>s of the hop. The chemistry of the hop resins <strong>and</strong> essential<br />
oils is discussed in detail in Chapter 8but it is useful to note here that from the brewing<br />
st<strong>and</strong>point the most important hop resins are the alpha-acids ( -acids) sometimes referred<br />
to as `alpha'. In conventional brewing hops are boiled with sweet wort for 1 1<br />
2<br />
2 hours<br />
during which time the -acids go into solution <strong>and</strong> are isomerized into the iso- -acids,<br />
the main bitter principles of beer. In open coppers the bulk of the essential oil constituents<br />
are vaporized during this period of boiling so brewers may add a portion of choice<br />
`aroma' hops late in the boil to replace this loss. Alternatively, dry hops may be added to<br />
beer, either in cask or conditioning tanks, to introduce hop aroma ± a process known as<br />
dry hopping.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 7.1 World production of hops <strong>and</strong> alpha acids ± 2002<br />
Country Hop acreage Production Alpha production<br />
(hectares) (metric tonnes) (metric tonnes)<br />
Australia 862 2384 316.9<br />
Austria 225 296 20.6<br />
Belgium 250 438 43.0<br />
Bulgaria 239 303 28.9<br />
China 5642 14167 1423.1<br />
Czech Republic 5968 6200 10.8<br />
Germany 18354 31500 2951.0<br />
France 814 1550 43.8<br />
New Zeal<strong>and</strong> 406 884 95.3<br />
Pol<strong>and</strong> 2197 1800 76.0<br />
Portugal 37 57 5.5<br />
Russia 862 440 56.6<br />
Slovakia 350 350 12.2<br />
Slovenia 1816 2200 151.0<br />
South Africa 500 965 117.5<br />
Spain 730 133 13.3<br />
UK ± Engl<strong>and</strong> 1896 2653 242.0<br />
Ukraine 1778 1000 48.9<br />
USA 11864 25815 3140.0<br />
Yugoslavia 493 616 35.1<br />
2002 Totals 55058 93455 8810.0<br />
Source: November 2002 IHGC Report<br />
7.2 Botany<br />
Three species of Humulus are known, H. lupulus, H. japonicus <strong>and</strong> H. yunnanensis<br />
(Neve, 1991). H. lupulus is indigenous to <strong>and</strong> is cultivated in much of the Northern<br />
hemisphere between 35ë <strong>and</strong> 55ëNbut it is also cultivated in the Southern hemisphere in<br />
Australia, New Zeal<strong>and</strong> <strong>and</strong> South Africa. H. japonicus is widespread in China <strong>and</strong> Japan<br />
but it lacks lupulin gl<strong>and</strong>s <strong>and</strong> therefore brewing value; it is sometimes grown as an<br />
ornamental garden plant. Little is known about H. yunnanensis from southern China.<br />
Humulus <strong>and</strong> Cannabis are the only two genera in the family Cannabinaceae; some<br />
authorities classify them in the Moraceae. Cannabis is represented only by C. sativa L.<br />
(Indian hemp, hashish, marijuana). There are some chemical similarities between hops<br />
<strong>and</strong> hashish but the resins of the two species are distinct; those of hops provide the bitter<br />
principles of beer while those of Cannabis include the psychotomimetic drug,<br />
tetrahydrocannabinol. Cannabis <strong>and</strong> Humulus spp. have been grafted on to each other<br />
but the characteristic resins do not cross the grafts.<br />
The hop is aperennial climbing plant; the aerial part dies off in the autumn but the<br />
rootstockstaysinthesoil, sometimesformanyyears.Theplantneeds asupportup which<br />
to grow. In the wild, hops are found in hedgerows but for cultivation they are trained up<br />
strings attached to permanent wirework. In the spring the stem tissue in the upper part of<br />
the rootstock produces numerous buds from which many shoots develop. The farmer<br />
selects the strongest shoots <strong>and</strong> trains them clockwise up the strings. As the bines climb,<br />
young flowering shoots develop in the leaf axils ±the so-called `pin' stage -which then<br />
form the young female inflorescence with papillated stigmas ±the `burr' (Fig. 7.1). From<br />
this the strobiles or hop cones develop. The cones consist of a central strig with bracts <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
(b)<br />
(e)<br />
(e)<br />
Fruit<br />
(seed)<br />
(g)<br />
(c)<br />
1 cm<br />
Bract<br />
scar<br />
Lupulin<br />
gl<strong>and</strong>s<br />
bracteoles attached. Most of the lupulin gl<strong>and</strong>s are formed at the base of the bracteoles<br />
but they are readily detached <strong>and</strong> adhere to the bracts, strig <strong>and</strong> seed (Fig. 7.1). A few<br />
lupulin gl<strong>and</strong>s are found on the undersides of hop leaves but not enough to make these<br />
useful for brewing. The lupulin gl<strong>and</strong>s can contain as much as 57% of -acids <strong>and</strong> the<br />
sum of the ( ‡ )-acids is equal to 75 6% of the weight of the gl<strong>and</strong>. The ratio / can<br />
range from 0 to about 4. The amount of resin/gl<strong>and</strong> is fairly constant; the `high-alpha'<br />
varieties (see later) contain many more gl<strong>and</strong>s than the `low-alpha' varieties. It is<br />
predicted that the maximum lupulin content/cone that could be obtained by breeding is<br />
about 32% w/w which corresponds to a ( ‡ ) content of about 23% of the cone (Likens<br />
et al., 1978) New varieties with over 16% -acid have been bred.<br />
The hop is dioecious, male <strong>and</strong> female flowers are produced on different plants. Male<br />
flowers have five sepals <strong>and</strong> five anthers but since the flowers drop off after flowering<br />
any brewing value is lost. However, the male flowers produce pollen which can be<br />
carried long distances by the wind so any female plant in the vicinity will be fertilized<br />
<strong>and</strong> produce seeds at the base of the bracteoles. Despite many demonstrations that<br />
excellent lager beers can be produced with seeded hops, lager brewers do not like seeds<br />
so most varieties are grown `seedless'. In Europe this means that dried hops contain less<br />
than 2% w/w of seed; in the USA the limit is 3%. It was shown in Engl<strong>and</strong>, as long ago as<br />
(f)<br />
(d)<br />
1 cm<br />
0.5 cm 0.1 cm<br />
Stipular<br />
bract<br />
Bracteole<br />
Fig. 7.1 Hop (Humulus lupulus L.) (a) young shoot; (b) male flowers; (c) `pin', young flowering<br />
shoot developing in the leaf axils; (d) `burr', young female inflorescence with papillated stigmas;<br />
(e) part of axis (`strig') of cone; (f) single mature hop cone; (g) bracteole with seed <strong>and</strong> lupulin<br />
gl<strong>and</strong>; <strong>and</strong> (h) lupulin gl<strong>and</strong> (After Burgess (1964) <strong>and</strong> Neve (1991) by kind permission of Kluwer<br />
Academic Publishers).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
(f)<br />
(h)
1908, that the yield/acre was higher if the hops were fertilized so English growers were<br />
encouraged to plant male hops in their gardens <strong>and</strong> by now many are wild in the<br />
hedgerows. Fertilized hops may contain as much as 25% w/w of seed. In contrast, in<br />
Germany, except in breeding stations, male hops must be removed by law. In the USA<br />
most varieties are grown seedless but in some areas males are planted. For example, a<br />
commercial yield of the variety Fuggle could not be obtained in the absence of male hops<br />
The bulk of the English hop crop contains seeds in excess of the European limit but in<br />
isolated areas male hops have been removed <strong>and</strong> the crop grown seedless. The hop plant<br />
is usually diploid with 20 chromosomes but triploid plants have been bred which are very<br />
infertile <strong>and</strong> have a low level of seeds even when pollinated.<br />
The growth of hops is strongly influenced by the amount of daylight. They require at<br />
least 13 hours of daylight for vegetative growth to occur; with shorter periods the plant<br />
becomes dormant. The plant must produce 20±25 nodes before being ready to flower<br />
when the days shorten. However, flowering will be inhibited if the days are too long. In<br />
South Africa, where the daylength is marginally too short, artificial light has been used to<br />
delay flowering <strong>and</strong> so improve the yield. Hops have also been grown in Kenya but here<br />
artificial light is necessary to allow vegetative growth as well as to delay flowering. A<br />
period of illumination in the middle of the night is probably more efficient than delaying<br />
nightfall.<br />
7.3 Cultivation<br />
Setting up a hop garden or yard requires considerable capital <strong>and</strong> the necessary wirework<br />
is not readily adaptable to any other crop so, once a garden is established, farmers will<br />
continue to grow hops even if it is barely profitable. The wirework must support the<br />
weight of the crop, probably in adverse weather conditions, so the corner posts must be at<br />
least 15 cm (6 in.) in diameter but the intermediate posts, every third or fourth hill, need<br />
not be so robust. Wooden posts are usually used but sometimes concrete or steel posts are<br />
more readily available. Before the advent of mechanical picking the wirework in English<br />
gardens was 3.75 4.25 m (12.5 14 ft.) high. At harvest the bine was cut down from the<br />
overhead wirework <strong>and</strong> the hops were picked by h<strong>and</strong> in the garden into coarse woven<br />
sacks called pokes. Hop picking in Kent was the annual holiday of many from the East<br />
End of London; in the West Midl<strong>and</strong>s the pickers came from Birmingham. As the bine<br />
withered the nutrients therein returned to the rootstock. This was not possible with<br />
mechanical picking when the bine was cut down from the wirework <strong>and</strong> above the<br />
rootstock <strong>and</strong> transported in a trailer to a static picking machine Then, higher wirework<br />
was advantageous as in America (4.0 5.5 m, 13 18 ft.) <strong>and</strong> continental Europe<br />
(6.0 7.0 m, 20 23 ft.).<br />
In height-of-wirework trials at Wye College, Kent, the maximum yield of most<br />
varieties was with 5 m (16 ft.) wirework but some vigorous varieties showed increased<br />
yields at 5.5 m (18 ft.). As well as the height of the wirework, the layout <strong>and</strong> stringing<br />
patterns in hop gardens can show considerable regional variations. Nowadays the rows of<br />
plants are usually 2.8 3.2 m apart to allow tractors to pass freely. In Germany <strong>and</strong><br />
continental Europe there is often a spacing of 1.5 m between plants or hills with two<br />
wires/hill. In the USA the spacing between rows <strong>and</strong> plants is about 2.25 m with two<br />
strings to each plant. In Engl<strong>and</strong> <strong>practice</strong> differs between Kent <strong>and</strong> the West Midl<strong>and</strong>s. In<br />
Kent the spacing between rows <strong>and</strong> plants within rows is 2 m but with the umbrella<br />
system of stringing there are four strings/hill. In Worcester the distance between rows<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
may be as much as 3 m but within the rows the plants are only 1 m apart with two strings/<br />
hill. One-year old plants will be used to lay out a new garden.<br />
Apart from breeding new varieties, hops are propagated vegetatively from `setts'.<br />
Three techniques are used to produce setts: (i) hardwood `strap cuts' are taken from the<br />
base of the rootstock in winter <strong>and</strong> allowed to root, (ii) layering, <strong>and</strong> (iii) mist<br />
propagation. For layering, a bine is allowed to grow until it is taller than the distance to<br />
the next hill. It is then taken down, laid along the ground, covered with soil, <strong>and</strong> the top of<br />
the bine trained up the next hill. In the autumn the hardwood bine is uncovered <strong>and</strong> cut<br />
up, each piece with a node, <strong>and</strong> planted out in nursery beds. With mist propagation,<br />
young growing shoots are taken <strong>and</strong> cut up each piece with two nodes <strong>and</strong> two leaves,<br />
rooted in sterilized peat at 21 ëC under an intermittent mist-like spray controlled by an<br />
`electronic leaf'. Under these conditions the plant remains turgid <strong>and</strong> rooting takes 10±14<br />
days. After hardening off the plants are transferred to a nursery garden. It is obviously<br />
most important that only healthy plant material should be used to lay out a new garden. It<br />
should be free both of viruses <strong>and</strong> viroids. In order to reduce the risk of infection,<br />
commercial hop propagators are usually sited well away from the main hop growing<br />
areas. For example, in Engl<strong>and</strong>, they are in East Anglia, well away from Kent <strong>and</strong> the<br />
West Midl<strong>and</strong>s.<br />
When a garden is laid out a strong hook or peg is put into the ground near each hill to<br />
facilitate stringing. In Engl<strong>and</strong> this is carried out by a man on the ground with a long pole<br />
<strong>and</strong> a continuous length of coir string. In bygone days stringing was carried out by men<br />
on stilts. In continental Europe cut lengths of soft wire are used. In the USA <strong>and</strong> Australia<br />
cut lengths of string are used <strong>and</strong> attached to the upper wire by men in a tractor-drawn<br />
tower. As might be expected the largest yield/plant was obtained with the widest spacing<br />
but the highest yield/hectare was obtained with closer spacing. The angle of slope of the<br />
string also influences the yield. In Engl<strong>and</strong> the highest yields were obtained with a slope<br />
of 65ë but in Germany with a slope of 72±78ë.<br />
Traditionally, hop growing involved intensive cultivation but today, with the high cost<br />
of labour, `non-cultivation' is becoming increasingly popular. This involves controlling<br />
the weeds with herbicides (Paraquat in the autumn <strong>and</strong> Simazine in the spring). It was<br />
found that non-cultivation did less damage to the soil structure. As mentioned above,<br />
when the bines are about 0.5 m (18 in.) long they are trained clockwise up a string;<br />
usually two or three bines/string. Where practical, complete or partial self-training may<br />
be employed. Excess shoots may be cut off or removed with a chemical defoliant. The<br />
latter technique is also used to remove leaves <strong>and</strong> laterals from the lowest 1±3 m (3±6 ft.)<br />
of the bine. This helps to discourage mildew <strong>and</strong> red spider mites. The defoliants used<br />
include tar oil with sodium monochloroacetate. Paraquat <strong>and</strong> Diquat can be used when<br />
the bines reach the top wire but may damage younger plants.<br />
The hop is a deep-rooted plant which requires a good depth of soil, the pH of which<br />
should be kept above pH 6.5 by liming. The luxuriant growth of the plant makes heavy<br />
dem<strong>and</strong>s on soil nutrients which must be replaced. However, the present view is that<br />
earlier fertilizer regimens, for example, up to 225 kg N/ha, were excessive. It is<br />
recommended that, wherever possible, fertilizer treatments should be based on soil<br />
analyses <strong>and</strong> for nitrogen should not exceed 135 kg/ha. In Germany it is recommended<br />
that fertilizer treatments should be calculated from the amount of nutrient removed with<br />
the crop. The soil nitrate level should also be monitored. Hops contain up to 1.2% w/w of<br />
nitrate, which could significantly influence the level of nitrates in beer. Low levels of<br />
nitrates are desirable in beer because they can be reduced to nitrites, which can react with<br />
primary <strong>and</strong> secondary amines to produce carcinogenic N-nitrosoamines. It is also<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ecommended that phosphate <strong>and</strong> potash fertilizer treatments should be based on soil<br />
analyses <strong>and</strong> not exceed 300kg P2O5/ha <strong>and</strong> 450kg K2O/ha. With high residual levels no<br />
further treatment may be necessary. Soil analysis may also indicate the need for<br />
additional magnesium, 30±100kg/ha, to prevent the plants becoming chlorotic. Trace<br />
element deficiencies due to alack of zinc <strong>and</strong>/or boron have been observed in hops.<br />
In Engl<strong>and</strong> <strong>and</strong> western Europe the water requirements of the hop crop are usually<br />
supplied by natural rainfall but elsewhere irrigation is often necessary. In the USA the<br />
crop requires 400±500mm (18±20in.) of rain in the Willamette valley, Oregon <strong>and</strong><br />
760mm (30in.) in the Yakima valley, Washington State, which is supplied by furrow<br />
irrigation. Elsewhere, in Australia <strong>and</strong> in the Backa area of Serbia, overhead sprinkler<br />
systems are used. It is claimed that irrigation produces better hops, of more even quality,<br />
than natural rainfall.<br />
As the hop ripens in the Northern hemisphere, the first traces of resin can be detected<br />
in early August, the -acids afew days before the -acids, <strong>and</strong> resin synthesis is almost<br />
complete by the end of the month. De Keukeleire et al. (2003) have measured the<br />
formation <strong>and</strong> accumulation of -acids, -acids, desmethylxanthohumol <strong>and</strong> xanthohumol<br />
during flowering. Essential oil synthesis starts later <strong>and</strong> in some varieties resin<br />
synthesis may be complete before essential oil synthesis starts. Oxygenated compounds<br />
<strong>and</strong> sesquiterpenes are formed first but as the hop ripens the synthesis of myrcene<br />
becomes quantitativelythe most important process.Different varieties mature at different<br />
rates so agrower may choose amixture of early <strong>and</strong> late maturing varieties to spread<br />
picking over three to four weeks in September.<br />
The characteristics by which agrower decides that the crop is ready for picking are<br />
(Burgess, 1964):<br />
1. The bracts <strong>and</strong> bracteoles close towards the axis of the cone giving it acompact<br />
form.<br />
2. Thefullgrowthoftheterminalbracteole,whenseeded,causesittoprotrudefromthe<br />
top of the cone.<br />
3. The bracts <strong>and</strong> bracteoles become firm <strong>and</strong> slightly resilient. They rustle when<br />
squeezed in the h<strong>and</strong> <strong>and</strong> are rather easily detached from the axis.<br />
4. Thecolourofthebracteoles<strong>and</strong>,toalesserextent,thebractschangestoayellowishgreen.<br />
5. The contents of the seed become firm. The fruit coat (pericarp) becomes brittle <strong>and</strong><br />
of apurplish colour.<br />
6. The lupulin gl<strong>and</strong>s are completely filled with resins.<br />
7. The aroma of the hop is fully developed.<br />
Hops should be picked as soon as possible after they become ripe; overripe cones tend to<br />
open <strong>and</strong> become more fragile <strong>and</strong> thus may be easily shattered by the wind, birds, or<br />
during picking. In all cases hops should be picked within ten days of ripening.<br />
As mentioned above, the bines are cut down from the top of the wirework <strong>and</strong> 1.2±<br />
1.5m(4±5ft.) from the ground <strong>and</strong> laid on atrailer for transport to astatic picking<br />
machine (Fig. 7.2). Here the bine is attached to atrackway <strong>and</strong>, depending on the design,<br />
enters the machine either horizontally or vertically. The hops <strong>and</strong> leaves are stripped from<br />
the bine by numerous moving wire hooks <strong>and</strong> then passed over various screens to<br />
separate the hop cones from unwanted debris. According to EEC regulations certified<br />
hops must not contain more than 6% of leaf <strong>and</strong> more than 3% of waste. The waste from<br />
the picking machine may be composted but should be burnt if there is any risk of disease.<br />
Mobile picking machines that can pick the hops in the garden have been designed but for<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Bine<br />
conveyor<br />
track<br />
Comb for<br />
disentangling<br />
bine<br />
Cones removed<br />
by flailing rotors<br />
Chopper for<br />
cutting up<br />
stripped bine<br />
Conveyor for<br />
cones, leaves<br />
<strong>and</strong> sprays<br />
removed by<br />
flails<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Revolving<br />
rollers for<br />
separating<br />
leaves<br />
Spray<br />
conveyor<br />
chain<br />
Cones<br />
removed<br />
from<br />
sprays<br />
Loose petals<br />
separated from<br />
leaves here <strong>and</strong><br />
may be conveyed<br />
to final hop harvest<br />
Leaves<br />
separated<br />
from cones<br />
Main waste<br />
lateral<br />
conveyor<br />
Fig. 7.2 Hop-picking machine (Courtesy of Bruff Mfg. Co. Ltd.)<br />
Screener to<br />
separate small<br />
leaves, pieces<br />
of stem from<br />
the hops<br />
Lateral waste<br />
conveyor<br />
Supply of<br />
cones<br />
to bags<br />
Bagging<br />
off point
high wirework they are large, heavy <strong>and</strong> clumsy <strong>and</strong> have not been widely used.<br />
However, mobile picking machines have been successfully used with dwarf hops (see<br />
pages 249±50). Green hops contain about 80% w/w moisture <strong>and</strong> must be dried as soon as<br />
possible after picking. While waiting to be put on the kiln the hops must not be allowed to<br />
`sweat' as this will seriously reduce the quality of the crop. Although the vast majority of<br />
the crop is dried, some brewers in hop growing areas make seasonal brews using green or<br />
partially dried hops (feathered at 40% moisture) which have exaggerated hop flavours.<br />
7.4 Drying<br />
Most hops are dried on the farm to a final moisture level of about 10% w/w or less.<br />
However, in Germany the farmer dries his hops to about 14% moisture <strong>and</strong> then sends<br />
them, loosely packed, to a merchant. The merchants sorts <strong>and</strong> blends the hops <strong>and</strong> then<br />
completes the drying to the specified level. In a traditional English circular oast house the<br />
hops were spread on a horsehair cloth on a slatted floor 4.5 m (12±16 ft.) above a charcoal<br />
or anthracite fire. The drying floor is equipped with two sets of doors, the green hops are<br />
loaded on one side <strong>and</strong> the dried hops removed onto a conditioning floor on the other.<br />
Only natural draught, produced by tapering the roof to a cowl, was available <strong>and</strong> this was<br />
less than 0.1 m/s (19 ft/min) <strong>and</strong> only capable of drying a shallow (20±27 cm, 8±12 in.)<br />
bed of hops.<br />
Modern kilns are more likely to be rectangular, the air heated directly or indirectly with<br />
an oil burner <strong>and</strong> blown, or sucked, through the bed of hops with a powerful fan (Fig. 7.3).<br />
If an oil burner heats the air directly it is necessary to ensure complete combustion so the<br />
hops are not contaminated with unburnt oil. To avoid this a heat exchanger or stove may be<br />
used to heat the air. The air speed <strong>and</strong> temperature have to be carefully controlled<br />
throughout the drying as -acids are destroyed with increasing air temperature. When the<br />
hops are first put on the kiln with c. 80% moisture, evaporation cools them below the air<br />
temperature but as the hops dry they approach the air temperature. If the air speed is too<br />
low, the air may become saturated with moisture in the lower layers of the bed <strong>and</strong> then<br />
deposit moisture on the hops in the upper layer. This `reek' results in serious discoloration<br />
of the hops which will result in a poor h<strong>and</strong> evaluation. On the other h<strong>and</strong>, air speeds of<br />
STORE FOR BAGS<br />
OF GREEN HOPS<br />
Loading door<br />
Slide to retain hops<br />
Open slatted floor<br />
10 feet<br />
KILN<br />
Hops<br />
Thermometer<br />
bulb<br />
Sulphur<br />
pan Fan<br />
Louvres<br />
COOLING ROOM<br />
Unloading door<br />
Press<br />
Slide to retain hops<br />
Thermometer dial<br />
Combustion Pocket<br />
chamber<br />
Oil burner<br />
Fig. 7.3 Modern oast house (Burgess, 1964)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Door<br />
Door
above 0.3 m/s (60 ft./min) are likely to blow hops off the bed (sometimes a wire mesh is<br />
used to retain them). In <strong>practice</strong> hops are usually dried with air between 60±80 ëC (140±<br />
176 ëF) <strong>and</strong> at a speed of 0.2±0.3 m/s (40±60 ft./min).<br />
For the most efficient use of fuel, the air leaving the bed of hops will be almost<br />
saturated with moisture. This is easily achieved when the hops are first put on the kiln but<br />
the removal of the last few percent of moisture is much less efficient. This does not<br />
matter where fuel oil is cheap, as in the USA where deep layers (up to 1 m, 40 in.) of hops<br />
are dried on large floors (e.g. 13.4 m 13.4 m, 32 ft. 32 ft.) at 60±65 ëC (140±150 ëF).<br />
In Europe, where fuel is more expensive, multi-layered kilns have been built with<br />
movable floors like venetian blinds. Green hops are put onto the top floor <strong>and</strong> when dried<br />
hops are removed from the bottom floor, the partially dried hops are dropped to a lower<br />
level <strong>and</strong> the top floor reloaded. Diagrams of such kilns are given by Neve (1991).<br />
However, such kilns are expensive to construct <strong>and</strong> a cheaper alternative with a single<br />
floor is to recirculate warm air towards the end of drying. In another method green hops<br />
are collected in large bins with wire mesh bottoms <strong>and</strong> the bins moved over hot air ducts<br />
at different temperatures. Continuous hop dryers have been built in Europe <strong>and</strong> again,<br />
diagrams are given by Neve (1991).<br />
It used to be normal English <strong>practice</strong> to burn sulphur in the kiln during the first 30±<br />
45 min. of drying. This caused the hops to assume a uniform bright yellow colour which<br />
was highly rated on h<strong>and</strong> evaluation. However, sulphuring has been shown to destroy -<br />
acids so the <strong>practice</strong> has been generally discontinued in Engl<strong>and</strong>. In Germany merchants<br />
often burn sulphur in the final drying of the hops <strong>and</strong> in the USA some hops, but not all,<br />
are sulphured. The experience of the oast man in determining when the drying is<br />
complete is important. The remaining moisture is unevenly distributed making sampling<br />
for analysis difficult. Within the bed of hops there is more moisture in the upper layers<br />
than in the lower <strong>and</strong> in the cone most moisture is found in the strig. Accordingly the<br />
hops are removed from the kiln on to a conditioning floor where they are covered with<br />
cloths <strong>and</strong> the moisture allowed to equilibrate for several hours.<br />
In Engl<strong>and</strong> hops are traditionally packed in jute/polypropylene sacks called pockets.<br />
They are usually about 2.1 m (7 ft.) long <strong>and</strong> 0.6 m (2 ft.) in diameter <strong>and</strong> hold about 76 kg<br />
(170 lb.) of hops. The empty pocket is suspended through a hole in the cooling room floor<br />
beneath a press with a circular foot, usually operated by an electric motor. The base of the<br />
suspended pocket is supported further by a strong canvas webbing belt. The cooled hops<br />
are pushed into the pocket with a canvas shovel called a scuppet. When the pocket is full of<br />
loosely packed hops the press is operated <strong>and</strong>, after compression, more hops are added.<br />
The process is repeated until the pocket is tightly packed with a density of 137±145 kg/m 3<br />
(8.5±9 lb./ft. 3 ). The pocket is then supported on the webbing belt while it is sewn up. In the<br />
USA, <strong>and</strong> increasingly elsewhere, hops are packed in rectangular bales measuring 137<br />
51 76 cm (4 ft. 6 in. 1 ft. 8 in. 2 ft. 6 in.). The baling press is essentially a steel box<br />
with detachable sides in which a ram operates. The box is lined with hessian <strong>and</strong> after the<br />
final filling a hessian cloth is placed on top. While the hops are compressed by the ram, the<br />
sides are removed <strong>and</strong> the hessian cloths sewed together. Bales are more expensive to<br />
produce than pockets but the hops are more densely packed (180 kg/m 3 , 13 lb./ft. 3 ) so they<br />
can be transported more efficiently. However, pockets can be rolled whereas bales have to<br />
be lifted. Before the advent of pellets bales were sometimes compressed to half their<br />
normal size for export. Most of the lupulin gl<strong>and</strong>s were ruptured by this treatment.<br />
Hops deteriorate on storage, in some cases significantly, before the next season's crop<br />
makes up 100% of the hop grist (70±100 weeks). The deterioration can be slowed either<br />
by cold storage at 0.20 ëC (33 ëF) <strong>and</strong>/or by storage in an inert atmosphere. For pockets or<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ales cold storage is the most practical but pellets are usually packed in an inert<br />
atmosphere. The pattern of deterioration shows alag period, which appears to be a<br />
varietal characteristic, probably due to the amount of antioxidants present (Lemusieau et<br />
al., 2001), followed by aperiod when the loss of both -<strong>and</strong> -acids can be fitted to<br />
either zero-order or first-order rate equations (Green, 1978). The chemical changes that<br />
take place during storage <strong>and</strong> the methods of chemical analysis are discussed in Chapter<br />
8. Samples for chemical analysis are taken with a`cork borer' sampler 70±80mm in<br />
diameter <strong>and</strong> 200±250mm long with a serrated edge at the cutting end. For h<strong>and</strong><br />
evaluation asquare sample approximately 10cm 10cm 10cm (4in. 4in. 4in.)<br />
is cut from the middle of the pocket. From this sample the expert will verify that it is true<br />
to type <strong>and</strong> variety, assess the wholeness of the cones on the face <strong>and</strong> cut side of the<br />
sample <strong>and</strong> assess the aroma by rubbing the cones between the palms <strong>and</strong> inhaling. The<br />
stickiness of the sample gives an indication of the resin content. The expert will also<br />
determine whether any cone damage or discoloration is due to superficial mechanical<br />
damage or wind bruising as opposed to pests <strong>and</strong> diseases. In particular he will look for<br />
aphid <strong>and</strong>/or red spider mite infestation, powdery mildew, downy mildew or other<br />
diseases. Further, he will assess that the leaf <strong>and</strong> strig content is below the EEC limit of<br />
6%, that the moisture content is below the EEC limit of 12% <strong>and</strong>, from the brightness or<br />
dullness of the sample, determine if the hops had been picked under wet conditions or if<br />
the air flow in the oast had not been properly controlled. English hops are graded on the<br />
basis of h<strong>and</strong> evaluation. For alpha/aroma varieties (see pages 249±52) <strong>and</strong> high alpha<br />
varieties there are two grades: grade Ihops attract the base contract price but there is a<br />
deductionforgradeIIhops.Foraromahopsthereisalsoa`choicest'gradethatreceivesa<br />
premium above the base contract price.<br />
7.5 Hop products<br />
Whole hops are abulky, sticky product not suited to automated delivery into the copper.<br />
They only contain about 20% of useful brewing materials which are concentrated in the<br />
lupulingl<strong>and</strong>s.Anyconcentrationoftheseactiveprincipleswillreducetransport<strong>and</strong>cold<br />
storage costs <strong>and</strong> give aproduct easier to store in an inert atmosphere. The main hop<br />
products are listed in Table 7.2. The non-isomerized products, added to the copper, are<br />
utilized (see page 271) no better than whole hops. Hop powders, although much denser<br />
than whole hops, are still sticky <strong>and</strong> unsuitable for automated dosing. They are not<br />
commercially available but are converted into pellets.<br />
7.5.1 Hop pellets<br />
For the preparation of hop pellets it may be necessary to dry the hops further so they<br />
contain 8±10% moisture. They are then cooled to 30 ëC <strong>and</strong> crushed in a hammer mill.<br />
For Type 90 pellets (normal hop pellets), the resulting powder (1 5 mm) is homogenized<br />
in an orbital screw mixer <strong>and</strong> then pelleted in a ring or horizontal die. Friction within the<br />
die will raise the temperature which should not be allowed to exceed 55 ëC. In the die the<br />
lupulin gl<strong>and</strong>s will be crushed so it is important that the pellets should be cooled <strong>and</strong><br />
packed as soon as possible. The pellets are usually packed in metallized polyethylene<br />
laminate foils (0.1 0.15 mm thick) either under a vacuum (hard packs) or under an inert<br />
gas (nitrogen or carbon dioxide) at atmospheric pressure (soft packs). The packs are<br />
usually protected in cardboard boxes <strong>and</strong> are ideally stored at 1 5 ëC. Pack sizes range<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 7.2 Hop products<br />
Non-isomerized<br />
Double-compressed whole hops<br />
Hop pellets Type 90<br />
Hop pellets Type 45<br />
Stabilized hop pellets<br />
Solvent extracts: hexane<br />
ethanol<br />
liquid carbon dioxide<br />
supercritical carbon dioxide<br />
Isomerized hop products<br />
Isomerised hop pellets<br />
Isomerized kettle extract<br />
Isomerized hop extracts for post-fermentation bittering<br />
Reduced isomerized hop extracts: Dihydro- (rho)-iso- -acids<br />
Tetrahydroiso- -acids<br />
Hexahydroiso- -acids<br />
Hop oil products<br />
Hop pellets Type 100<br />
Oil-rich hop extract<br />
Pure hop oil: steam distillation<br />
molecular distillation<br />
Hop oil emulsions<br />
Fractionated hop oil<br />
Dry hop essences<br />
Late hop essences ± spicy, floral, estery <strong>and</strong> citrussy<br />
Miscellaneous<br />
Base hop fraction<br />
Purified beta fraction<br />
(European Brewery Convention Manual of Good Practice ± Hops <strong>and</strong> Hop Products, 1997)<br />
from 2 150 kg or may contain a specific weight of -acid so that the whole contents of a<br />
pack may be added to the copper without further weighing. The pellets are approximately<br />
6 mm 10 15 mm with a bulk density of 480 550 kg/m 3 (cf. whole hops at c. 140 kg/<br />
m 3 ). Such pellets are called Type 90 because roughly 90 g of pellets are obtained from<br />
100 g of hops but 98% of the -acids (dry weight) are recovered. At the start of the 21st<br />
century over 50% of the hop crop was processed into pellets. At least one large hop farm<br />
in the USA processes the hop crop directly into pellets.<br />
For lupulin-enriched hop pellets (Type 45) the hop powder from the hammer mill is<br />
sieved at 30 ëC. The material that passes through a 0.3 mm sieve contains the lupulin<br />
gl<strong>and</strong>s <strong>and</strong> represents about half of the weight of hops used; the larger particles go to<br />
waste. The bulk density of the Type 45 pellets is similar to that of the Type 90 pellets but<br />
the Type 45 pellets contain twice as much of the brewing principles. By further sieving it<br />
is possible to obtain a fraction composed almost entirely of lupulin gl<strong>and</strong>s but this is not<br />
necessary for brewing purposes.<br />
Stabilized hop pellets are prepared by mixing up to 2% by weight of magnesium (or<br />
calcium) oxide with the hop powder before pelletization. In the die the -acids are<br />
converted into their salts which are more stable than the free acids. During storage the -<br />
acid salts may isomerise into iso- -acid salts which are better utilized than -acid salts. If<br />
stabilized hop pellets (in soft packs) are kept at 45±55 ëC for 10±14 days, the<br />
isomerization is complete <strong>and</strong> isomerized hop pellets are formed.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
7.5.2 Hop extracts<br />
Many different solvents have been used to extract the brewing principles from hops but<br />
brewers have become increasingly worried about the possibility of solvent residues in<br />
their beer so only hexane (b.p. 69ëC), ethanol (b.p. 78ëC), <strong>and</strong> carbon dioxide are still<br />
used <strong>and</strong> of these carbon dioxide is the most important. Although hexane is widely used<br />
in the food industry, the only hop extract plant using hexane is scheduled to close.<br />
Hexane is anon-polar solvent which will only extract soft resins <strong>and</strong> hop oil constituents<br />
from hops. In contrast, ethanol is miscible with water so an ethanol extract will also<br />
contain hard resins <strong>and</strong> polyphenols. After removal of the ethanol amixture of resins <strong>and</strong><br />
ahot water extract is obtained but now only the resin extract is normally used as the hot<br />
water extract is rich in nitrates.<br />
At room temperature carbon dioxide is agas so it can only be used as asolvent under<br />
pressure. The phase diagram (Fig. 7.4) shows that both liquid carbon dioxide (below<br />
31ëC <strong>and</strong> 73 bar abs) <strong>and</strong> supercritical CO2 (above 31ëC <strong>and</strong> 73 bar abs) can be used for<br />
extraction. Liquid carbon dioxide has been used in Engl<strong>and</strong> <strong>and</strong> Australia but elsewhere<br />
supercriticalcarbondioxide hasbeen thechoice.Despite claims madetothecontrary,the<br />
composition of the two extracts is very similar. The liquid CO2 extract is pale yellow <strong>and</strong><br />
may contain fewer hard resins <strong>and</strong> polar bitter substances than the supercritical CO2<br />
extracts which are yellow to green in colour. Both are free of polyphenols <strong>and</strong> nitrates<br />
<strong>and</strong> contain few pesticide residues. Extraction with both liquid <strong>and</strong> supercritical CO 2is a<br />
batch process. Figure 7.5 is adiagram of aliquid CO 2 extraction plant. The extractor is<br />
charged with remilled hop pellets <strong>and</strong>, typically, carbon dioxide at 60±65 bar <strong>and</strong> 5±15 ëC<br />
is passed through the powdered pellets. The carbon dioxide is then evaporated from the<br />
extract at 40±50 ëC, condensed <strong>and</strong> returned to the extraction cycle. For supercritical CO2<br />
extraction the cooler in Fig 7.5 is replaced by a heat exchanger. Liquid CO2 at 60±70 bar<br />
is raised to the extraction pressure (200±250 bar) by a pump <strong>and</strong> the supercritical liquid<br />
raised to 40±60 ëC by a heat exchanger. After extraction the pressure on the extract is<br />
reduced to 60±80 bar before evaporation. The CO 2 is recovered, condensed <strong>and</strong> recycled<br />
as with liquid CO 2 extraction. Such CO 2 extracts can be used in the copper or as the raw<br />
material for isomerized extracts or molecular distilled oils.<br />
Pressure (bar abs)<br />
140<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
Solid<br />
phase<br />
Triple point<br />
5.18 bar abs<br />
–56.6°C<br />
Liquid phase<br />
Liquid extraction<br />
Vapour<br />
phase<br />
Supercritical<br />
phase<br />
Supercritical<br />
extraction<br />
Critical point<br />
73 bar abs<br />
31°C<br />
0<br />
–100 –50 0<br />
Temperature (°C)<br />
50 100<br />
Fig. 7.4 Pressure temperature equilibria for carbon dioxide (Benitez et al., 1997).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Extractor 1<br />
Cooler<br />
Pump<br />
Extractor 2<br />
Vapour phase<br />
carbon dioxide<br />
Condenser<br />
Extract<br />
Liquid phase carbon dioxide<br />
Evaporator<br />
Separator<br />
Liquid carbon<br />
dioxide buffer<br />
Fig. 7.5 Liquid carbon dioxide extraction with two extractors (Benitez et al., 1997).<br />
Hop extracts are usually packed in cans with food-grade linings. The cans can contain<br />
0.5±5.0 litres or aspecified weight of -acids. The use of glucose syrup or ahot water<br />
extract of hops to dilute resin extracts is now thought undesirable; such diluted extracts<br />
have amuch shorter shelf-life. For use, holes are punched in the cans which are then<br />
suspended (in abasket) in boiling wort. On alarger scale the extract is warmed to<br />
increase its mobility <strong>and</strong> then pumped into the copper (carbon dioxide extracts are less<br />
viscous than those prepared with hexane or ethanol). Isomerized <strong>and</strong> reduced hop<br />
extracts, designed for post-fermentation addition, are discussed in Chapter 8.<br />
7.5.3 Hop oils<br />
By definition, essential oils are volatile in steam so they are prepared by steam<br />
distillation. Usually minced hops are boiled with water <strong>and</strong> the condensate is collected in<br />
atrap which retains the oil <strong>and</strong> allows the aqueous phase to return to the boiler ±a<br />
process known as cohobation (the same technique is used in the analytical determination<br />
of hop oil). Such hop oil preparations have been available for many years, <strong>and</strong> some<br />
brewers used them in place of dry hopping, but they do not retain the true aroma of the<br />
hop <strong>and</strong> beers so treated could usually be distinguished from those which had been dry<br />
hopped. It is likely that some constituents are damaged at 100ëC <strong>and</strong> any constituents<br />
slightly soluble in water would be washed out in the trap.<br />
By steam distillation under reduced pressure (0.008mm Hg) at 25ëC, Pickett et al.<br />
(1977) obtained hop oil emulsions that were comparatively stable <strong>and</strong> imparted asound<br />
hop character to beer. However, hop oil constituents are soluble in liquid <strong>and</strong> supercritical<br />
CO2 with no risk of thermal degradation. The hop oil constituents are more soluble in<br />
liquid CO2 than the resins so extraction of a column of milled hops with 10±15% of the<br />
liquid CO2 needed for complete extraction will recover most of the hop oil with little of<br />
the resins. Similar preparations can be obtained with supercritical CO 2 but in <strong>practice</strong> it is<br />
better to aim for complete extraction. The pressure on the extract is then reduced to 100±<br />
120 bar when the hop acids are precipitated; the hop oil constituents stay in solution <strong>and</strong><br />
can be recovered by evaporation. Hop oils can also be recovered from CO2 extracts by<br />
molecular distillation at low pressure (0.001 mm Hg). Such oils prepared from a single<br />
cultivar of hops retain the characteristic aroma of that cultivar <strong>and</strong> can be used, with a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
food grade emulsifier, to produce hop oil emulsions (normally with an oil content of<br />
0.25% v/v). The production of dry hop essences is discussed in Chapter 8.<br />
7.6 Pests <strong>and</strong> diseases<br />
Today many consumers are prepared to pay a premium for foodstuffs produced<br />
`organically', that is without the use of agricultural chemicals. To meet this market<br />
some brewers are producing `organic' beers from `organic' barley <strong>and</strong> hops. However,<br />
hops are susceptible to attack by many pests <strong>and</strong> diseases so most hop growers need to<br />
use agrochemicals to produce acommercial crop. The agrochemicals which can be used<br />
are licensed by national or international bodies who also set amaximum allowable<br />
residue (MRL, mg/kg) for the chemical in the final product. To complicate matters the<br />
infecting organisms may develop resistance to the agent used so new agents have to be<br />
developed <strong>and</strong> approved periodically. Further, different strains of adisease may be<br />
susceptible to different agents. In general, more pests <strong>and</strong> diseases are found in the<br />
long-established hop-growing areas of Europe <strong>and</strong> North America than in the Southern<br />
hemisphere.<br />
7.6.1 Damson-hop aphid (Phorodon humuli Schrank)<br />
The most serious pest in the Northern hemisphere is the damson-hop aphid (Fig. 7.6)<br />
which overwinters as shiny black eggs in the bark of Prunus spp. (damson, sloe or<br />
plum). In early April wingless female insects hatch out <strong>and</strong> give birth to live young<br />
which rapidly multiply. After several generations winged females (alatae) arise which<br />
migrate to the hop when the flight threshold temperature of 13 ëC (55 ëF) is reached in<br />
late May or early June. The migrating alatae may not colonize the first hop plant on<br />
which they l<strong>and</strong> <strong>and</strong> may show a varietal preference. However, they feed <strong>and</strong><br />
reproduce mainly on leaves on the top of the bine. They insert their long stylets into<br />
the phloem str<strong>and</strong>s of the leaves to feed which weakens the leaves <strong>and</strong> causes<br />
defoliation. The most serious situation is when the alatae enter the cones. In addition<br />
the honeydew that the aphids produce supports the growth of sooty moulds, which<br />
lower the value of the hops on h<strong>and</strong> evaluation. More heavily infected cones will turn<br />
brown <strong>and</strong> limp <strong>and</strong> will probably shatter during machine picking. Those hops that<br />
survive will not find a ready market. Any aphids that survive until September±<br />
October, when the light period falls below 13.5 h/day, form winged females which<br />
Fig. 7.6 Damson-hop aphid (Phorodon humuli Schr.).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a) Insecticides/Acaricides<br />
R1<br />
R2<br />
F3C<br />
C C<br />
H<br />
H<br />
H3C CH3<br />
(7.1) R1 = R2 = Cl, R3 = H<br />
(7.2) R1 = R2 = Cl, R3 = F<br />
(7.3) R1 = R2 = Br, R3 = H<br />
CO2 C<br />
H<br />
(7.4) R1 = CF3, R2 = Cl, R3 = H<br />
Cl<br />
H3C<br />
H3C<br />
H3C CH3<br />
C C<br />
H<br />
H<br />
H3C CH3<br />
N<br />
CO2<br />
CO2<br />
CN<br />
C<br />
H<br />
CN<br />
H<br />
CH2<br />
migrate back to Prunus spp. The males follow later <strong>and</strong> the fertilized females lay the<br />
eggs which overwinter.<br />
Control includes removing Prunus sp. from nearby hedgerows but the alatae can fly<br />
considerable distances. It is important that the grower knows when migration onto the<br />
hop starts so that he can maximize his control measures. Usually this involves spraying<br />
with suitable insecticides such as the synthetic pyrethoids. Cyfluthrin, Cypermethrin,<br />
Deltamethrin, Fenpropathrin, Lambda-cyhalothrin, or Imidacloprid (Fig. 7.7). For many<br />
years organo-phosphorus insecticides were used but in Europe most aphids have<br />
developed resistance to these agents but they may still be effective elsewhere. As<br />
mentioned, most brewers are not happy with the risk of agrochemical residues in their<br />
R3<br />
Cypermethrin<br />
Cyfluthrin<br />
Deltamethrin<br />
(7.5) Fenpropathrin<br />
H<br />
(7.6) Bifenthrin<br />
Cl CH2 N<br />
O<br />
Lambda-Cyhalothrin<br />
O<br />
CH3<br />
N·NO2<br />
(7.7) Imidacloprid (Admire)<br />
Fig. 7.7 (a) Insecticides/acaricides, (b) herbicides <strong>and</strong> (c) fungicides accepted for use on hops<br />
(British Beer <strong>and</strong> Pub Association, Technical Circular No. 376, 2003).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
NH
Fig. 7.7 Continued.<br />
raw materials, so there is considerable interest in the biological control of aphids using,<br />
for example, ladybirds (Coccinellidae) but such control is never complete. Hops with<br />
resistance to aphids have now been bred <strong>and</strong> are being evaluated. The ASBC describe<br />
methods for estimating the number of aphids in a sample of hops.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 7.7 Continued.<br />
7.6.2 (Red) Spider Mite (Tetranchus urticae Koch)<br />
The spider mite is widespread <strong>and</strong> flourishes in hot dry conditions. The bright red females<br />
overwinter in the soil, under leaves or in cracks in hop poles. In the spring they climb the<br />
bines <strong>and</strong> suck sap from epidermal <strong>and</strong> sub-epidermal cells. They lay small translucent<br />
eggs <strong>and</strong> the mites that emerge are greenish-yellow with black markings (two-spotted<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mite).Thefirstsignofmiteattackisasilveryspecklingofthehopleaves.Smallnumbers<br />
of mites do not do much damage but severe infestation may result in loss of crop. When<br />
hops were first treated with organo-phosphorous insecticides there was good control of<br />
both aphids <strong>and</strong> mites but later both species developed resistance. The acaricides in use<br />
today include Bifenthrin <strong>and</strong> Propargite.<br />
7.6.3 Other pests<br />
Atleast40insectspecieshave beenfound living onthehopbut,with theexceptionofthe<br />
species discussed above, they do not usually cause serious damage. Other pests include<br />
the dagger nematode (Xiphinema diversicaudatum), which may be the vector for the<br />
Arabis Mosaic Virus (AMV), the hop-root eelworm (Heterodera humuli), clay-coloured<br />
weevils (Otiorrhynchus singularis), the rosy rustic moth (Hydroecia micacea), the flea<br />
beetle (Psylliodes attenuata Koch), earwigs (Forficula auricularis), wireworms<br />
(Agricotes spp.), <strong>and</strong> slugs (Agriolimax reticulatus <strong>and</strong> Arion hortensis) (Neve, 1991).<br />
7.6.4 Downy Mildew (Pseudoperonospora humuli (Miyabe <strong>and</strong> Tak.) G. W.<br />
Wilson)<br />
Worldwide, downy mildew is probably the most serious disease of hops. It was first<br />
observed in Japan in 1905, in the USA in 1909, <strong>and</strong> in Engl<strong>and</strong> <strong>and</strong> Europe by 1920. It is<br />
found in most hop-growing areas of Europe <strong>and</strong> North America but, by applying strict<br />
quarantine precautions, has been kept out of Australia, New Zeal<strong>and</strong>, <strong>and</strong> South Africa.<br />
The fungus survives the winter in the rootstock of the hop plant. In spring, when the hop<br />
produces numerous shoots, it migrates to infect some of the shoots so that they do not<br />
develop <strong>and</strong> produce stunted basal spikes which are the characteristic symptom of the<br />
disease (Fig. 7.8). On the undersides of the leaves of these spikes numerous black spores<br />
(conidia) develop which readily infect other young leaves (Fig. 7.9). Under wet<br />
conditions the conidia germinate on the leaves <strong>and</strong> produce motile zoospores which enter<br />
the plant through open stomata. Such infection causes black angular spots on the leaves.<br />
Infection of the growing tip of the bine will cause extension to cease <strong>and</strong> the formation of<br />
a basal spike. If the zoospores infect the flower or `burr' no cone will develop. If they<br />
enter a developing cone serious losses will occur; some of the bracts <strong>and</strong> bracteoles will<br />
turn brown giving the cones a variegated appearance <strong>and</strong> lower the value of any crop.<br />
Control involves the early removal of all basal spikes <strong>and</strong> lower leaves with infecting<br />
conidia. If this is done by h<strong>and</strong> all infected material should be burnt but by spraying with<br />
defoliating chemicals (anthracene oils, Diquat, Paraquat, or sodium monochloroacetate)<br />
any spikes will be killed in situ. Before downy mildew appeared on hops, copper<br />
fungicides, such as Bordeaux mixture, were widely used to control mildew on grapes, <strong>and</strong><br />
such treatments were readily applied to hops. Later, copper oxychloride was found to be<br />
less phytotoxic than Bordeaux mixture. In the 1950s organic dithiocarbamates such as<br />
Zineb were favoured but the Pesticide Safety Directorate has revoked the approval of<br />
such chemicals. The most effective fungicides in use today are Metalaxyl (Ridomil) <strong>and</strong><br />
Fosetyl-aluminium (Aliette). These can be used as foliar sprays or as a soil drench when<br />
they are capable of eliminating the fungus from the rootstock. However, this latter<br />
method of application has been discontinued in Engl<strong>and</strong> in fear that the fungus was<br />
developing resistance to Metalaxyl. Metalaxyl is applied as a foliar spray in admixture<br />
with copper oxychloride. Other approved fungicides include Chlorothalonil, Fenpropimorph<br />
(Corbel), Myclobutanil, <strong>and</strong> Peconazole (Topas). Despite the fact that Metalaxyl<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Healthy<br />
shoot<br />
Earth<br />
Basal<br />
spike<br />
Fig. 7.8 Hop plant infected with downy mildew (Pseudoperonospora humuli). Left: healthy<br />
shoot. Right: basal spike.<br />
50 μm<br />
Fig. 7.9 Downy mildew (Pseudoperonospora humuli). Sporangiophore with sporangia (Burgess,<br />
1964).<br />
can eliminate the fungus from the rootstock, it is good <strong>practice</strong> to dig up <strong>and</strong> burn any<br />
rootstock that has been infected <strong>and</strong> replace it with healthy material. Varieties with<br />
resistance to downy mildew are being bred.<br />
7.6.5 Powdery mildew (Sphaerotheca macularis (DC.) burr)<br />
Mould, white mould, red mould or powdery mildew are synonyms for this fungal disease<br />
that appears as white pustules on the leaves. Towards the end of the season red<br />
overwintering spores (perithecia) are formed. The disease has little effect on the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
vegetative growth of the host but serious losses occur if the cones are infected. The<br />
disease is probably more troublesome when non-cultivation is practised <strong>and</strong> plant debris<br />
infected with perithecia is left on the ground. Control involves treatment with sulphur,<br />
before burr, Triadimefon (Bayleton) <strong>and</strong> Bupirimate (Nimrod). Triforine (Saprol), which<br />
may reduce the yield, is usually reserved for serious infections.<br />
7.6.6 Verticillium Wilt (Verticillium albo-atrum Reinke <strong>and</strong> Berth)<br />
This disease, first detected in the Weald of Kent in 1924, is serious in Engl<strong>and</strong> <strong>and</strong><br />
Germany. It was thought, at first, to exist in two forms: mild (`fluctuating') <strong>and</strong> severe<br />
(`progressive') but the present view is that there is a continuum of strains giving infection<br />
ranging from very mild to very severe. There is no simple test, chemical or biological, to<br />
distinguish between the strains except by their behaviour towards a susceptible cultivar,<br />
e.g., Fuggle. Such tests, whether outside or in growth chambers, are of necessity slow. The<br />
fungus occurs in the soil either as spores (condia) or as mycelium on infected debris (Fig.<br />
7.10). It enters the hop roots <strong>and</strong> the dark coffee-coloured mycelium spreads through the<br />
500 μm<br />
10 μm<br />
10 μm<br />
(c)<br />
(d)<br />
(b)<br />
(a)<br />
50 μm<br />
Fig. 7.10 Verticillium wilt (Verticillium albo-atrum). (a) conidiophore; conidial heads <strong>and</strong><br />
conidia; (b) groups of conidiophore; (c) `dark' mycelium; (d) conidia (Burgess, 1964).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
vascular system of the plant to the leaves. The infected leaves develop yellow patches <strong>and</strong><br />
black necrotic areas between the veins giving the so-called `tiger stripe' effect before they<br />
fall off. Another symptom is that the lower 1.2±1.5 m (4±5 ft.) of an infected bine becomes<br />
swollen <strong>and</strong> may become detached from the rootstock. The wood of the infected bine<br />
shows brown areas due to the mycelia. In `fluctuating' wilt the brown colour is restricted<br />
to the centre of the bine <strong>and</strong> the hill will probably survive with little or no increase in wilt<br />
the following year. In severe `progressive' wilt the symptoms appear earlier in the season<br />
<strong>and</strong> are readily transferred to other plants in the garden especially in the direction of<br />
cultivation (such observations helped to popularize non-cultivation). Infected plant debris<br />
is readily carried on the wind, on boots <strong>and</strong> machinery to spread the disease.<br />
At present no form of wilt responds to chemical agents so the only control measures<br />
are hygiene <strong>and</strong> the breeding of wilt-tolerant varieties. The Progressive Wilt of Hops<br />
Order (1947) attempted to restrict the disease to the Weald of Kent where wilt-tolerant<br />
varieties could be planted. Elsewhere, in `eradication' areas, only susceptible varieties<br />
could be grown <strong>and</strong> any outbreak of the disease had to be notified, the infected plants <strong>and</strong><br />
those nearby had to be grubbed <strong>and</strong> all infected materials burnt. The infected area was<br />
then fenced off <strong>and</strong> kept under grass for several years. Wilt-tolerant varieties must not be<br />
grown in `eradication' areas as they may be symptomless carriers. These strict measures<br />
managed to keep the disease in check in East Kent <strong>and</strong> Hampshire <strong>and</strong>, for a time, the<br />
West Midl<strong>and</strong>s. Later, wilt spread rapidly through the West Midl<strong>and</strong>s so wilt-tolerant<br />
varieties may now be planted there on clean l<strong>and</strong>. The wilt-tolerant varieties are discussed<br />
below. Other fungal diseases which infect hops from time to time include Fusarium<br />
canker (Fusarium sambucinum Fuckel), black root rot (Phytophthora citricola Sawaba),<br />
grey mould (Botrytis cinerea Pers.), Alternaria alternata, black mould (Cladosporum sp.)<br />
<strong>and</strong> armillaria root rot (Armillia mellea mellea (Fr.) Quel.).<br />
7.6.7 Virus diseases<br />
In the past, virus diseases caused considerable damage to hops. Nettlehead was perhaps<br />
the most serious, the symptoms of which were well known before viruses had been<br />
recognized. In 1966 Bock identified arabis mosaic virus (AMV) <strong>and</strong> necrotic ringspot<br />
virus (NRSV) in hops. All hop plants with nettlehead contained AMV but not all plants<br />
infected with AMV developed nettlehead. It was found that a satellite of low molecular<br />
weight nucleic acid (SNA) was necessary as well as AMV to produce nettlehead. It was<br />
also found that the soil-borne dagger nematode (Xiphinema diversicaudatum) was a<br />
vector for AMV. L<strong>and</strong> infected with nematodes should be fumigated with dichloropropene<br />
<strong>and</strong> left fallow for two years before replanting with hops.<br />
Hop mosaic virus (HMV) is a carlavirus transmitted by aphids feeding on one plant<br />
<strong>and</strong> then moving to another. In English Goldings it causes severe stunting, the bines fall<br />
away from the strings, <strong>and</strong> the leaves show translucent b<strong>and</strong>ing along the veins <strong>and</strong> curl<br />
downwards. Most other varieties are symptomless carriers of the virus <strong>and</strong> there is no<br />
evidence that HMV causes any reduction in yield with them. Nevertheless, Goldings<br />
should not be planted with, or adjacent to, other varieties. Other carlaviruses are the hop<br />
latent virus <strong>and</strong> the American hop latent virus. The Prunus necrotic ringspot virus<br />
(NRSV) rarely produces recognizable symptoms in hops but laboratory tests showed that<br />
virtually 100% of the older varieties were infected. Comparison of yields from virus-free<br />
<strong>and</strong> infected plants showed that the infected plants contained 30% less -acid.<br />
Enzyme-linked immunosorbent assay (ELISA) tests have greatly facilitated the<br />
detection <strong>and</strong> characterization of viruses in hops. Virus can be eliminated from hop plants<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
y a combination of heat treatment <strong>and</strong> meristem-tip culture. Multiplication of viruses is<br />
inhibited by maintaining the plants at high temperature <strong>and</strong> if the tip of the apical shoot is<br />
dissected out of a plant <strong>and</strong> grown under sterile conditions, the regenerated plants are<br />
usually virus-free. By such techniques virus free clones of most varieties are available to<br />
hop propagators so most new planting material, with A+ certification, is virus free.<br />
Viroids consist of a small piece of RNA without a protein coat so they cannot be<br />
detected serologically. At least two viroids have been detected in hops. The hop stunt<br />
viroid has been found only in Japan but the hop latent viroid was found to be widespread<br />
in Germany <strong>and</strong> Engl<strong>and</strong>; the English variety Omega was particularly susceptible.<br />
However, virus- <strong>and</strong> viroid-free planting material is now available for most cultivars.<br />
7.7 Hop varieties<br />
The first cultivators of hops selected the most vigorous wild plants they could find. If<br />
their neighbours' plants gave better yields they would seek planting material from them<br />
so that, in time, the same cultivar would be grown in a given area. As transport improved<br />
these areas grew larger. Indeed, German hops were named after the area where they grew,<br />
e.g., Hallertau, Tettnang, Spalt <strong>and</strong> Hersbrucker. Early growers only had small kilns or<br />
oasts to dry their crop so they were interested in having early <strong>and</strong> late varieties to<br />
lengthen the harvest period, e.g., Hallertau mittelfruÈh, Hersbrucker spaÈt.<br />
An early rhyme suggests that:<br />
Hops, Reformation, Bays <strong>and</strong> Beer<br />
Came to Engl<strong>and</strong> in One Bad Year<br />
<strong>and</strong> that year was 1524 when Flemish immigrants brought hops to Kent. However, the<br />
finding of hop residues in a boat found in Graveney Marsh <strong>and</strong> dated c. AD 949 suggests<br />
they were known earlier (Wilson, 1975). Many different selections were grown in<br />
Engl<strong>and</strong>. According to Burgess (1964) the Hops Marketing Board classified Amos's<br />
Early Bird (selected in 1887), Bramling (before 1865), Cobbs (1881), Eastwell Golding<br />
(before 1889), Petham Golding, Rodmersham Golding (1880), Mathon (1901), <strong>and</strong><br />
Canterbury Goldings as Goldings <strong>and</strong> Tutsham <strong>and</strong> Whitbread's Golding Variety (WGV<br />
or 1147) as Golding varieties. Before genetic fingerprinting it was not possible to confirm<br />
how similar these varieties were. British brewers preferred Goldings but they were<br />
susceptible to mosaic viruses <strong>and</strong> the variety introduced by Richard Fuggle in 1875,<br />
which bears his name, was not. So Fuggles became popular with growers <strong>and</strong>, to a lesser<br />
extent, with brewers <strong>and</strong> in 1949 made up 78% of the English hop acreage. Goldings<br />
contain slightly more -acid than Fuggles <strong>and</strong> have a better aroma; Fuggles was used in<br />
the copper <strong>and</strong> Goldings for dry hopping. Brewers were usually prepared to pay a<br />
premium for Goldings.<br />
The early settlers introduced hop growing along the eastern seaboard of North<br />
America but from 1900 it moved to the North-West States of Washington, Oregon,<br />
California <strong>and</strong> Idaho. The main variety grown was Clusters (Early <strong>and</strong> Late) which<br />
contained more resin than Goldings or Fuggle but had a strong aroma described as<br />
`blackcurrant' or `tom-cat' which was disliked by most British brewers. Nevertheless<br />
some British brewers imported Clusters, <strong>and</strong> Salmon at Wye College set out to breed<br />
hops with more resin but English aromas. From an open pollinated wild hop from<br />
Manitoba (BB1) he obtained Brewer's Gold <strong>and</strong> Bullion which gave excellent yields <strong>and</strong><br />
were richer in -acids than other varieties then available. A cross between a male<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
seedling of Brewer's Gold <strong>and</strong> a Canterbury Golding gave Northern Brewer. These three<br />
new varieties were planted worldwide, in particular Bullion was grown in Oregon <strong>and</strong><br />
Brewer's Gold in Germany.<br />
Later, Northern Brewer became important in Germany because of its resistance to the<br />
German strain of Verticillium wilt <strong>and</strong> by 1978 these three varieties accounted for 47% of<br />
the German hop acreage. They were also grown in Belgium, Bulgaria, Spain <strong>and</strong> what<br />
was East Germany. They were less popular in Britain because of their `American' aroma.<br />
The Germans also maintained that their traditional varieties had better aromas than the<br />
new varieties <strong>and</strong> this led to the classification of hops as either `Aroma' hops or `Bitter'<br />
hops. The latter usually had higher -acid contents <strong>and</strong> were put into the copper at the<br />
beginning of the boil when any unpleasant volatiles would be evaporated. The `Aroma'<br />
hops were reserved for late addition or dry hopping <strong>and</strong> usually comm<strong>and</strong>ed a higher<br />
price than `Bitter' hops. However, there is no reason why a high -acid hop should not<br />
have a good aroma. Salmon bred many more new varieties of hops, some of which were<br />
grown commercially for a time encouraged by the Association of Growers of New<br />
Varieties of Hops (1944±1998). They include Concord, Brewers' St<strong>and</strong>by, Early Choice,<br />
Copper Hop, Quality, College Cluster, Brewers' Favourite, Sunshine, Malling<br />
Midseason, <strong>and</strong> Norton Court Golding. These, <strong>and</strong> many only assigned code numbers,<br />
served as the parents of the next generation of new varieties.<br />
The devastation caused by Verticillium Wilt in Kent prompted an urgent need for wilttolerant<br />
varieties. From among Salmon's seedlings Keyworth's Early, Keyworth's<br />
Midseason <strong>and</strong> Bramling Cross were found to be wilt tolerant but they were never very<br />
popular. The Whitbread Golding Variety was also wilt tolerant <strong>and</strong> was more acceptable.<br />
However, conservative brewers wanted a wilt-tolerant replacement for Fuggle <strong>and</strong><br />
Density, Defender, <strong>and</strong> Janus were introduced in 1959 to meet this need. Again they were<br />
not very popular but Progress <strong>and</strong> Alliance, introduced in 1966, were more acceptable<br />
<strong>and</strong> Progress is still grown today. However, by this time there had been many brewery<br />
amalgamations <strong>and</strong> the accountants had more say in the boardrooms. They knew that -<br />
acids, which could be measured, produced bitterness so they wanted high-alpha hops to<br />
produce bitterness as cheaply as possible. Aroma, which could not be easily measured,<br />
was much more subjective. Wye Northdown <strong>and</strong> Wye Challenger, introduced in 1971,<br />
had higher levels of -acid <strong>and</strong> good aroma but were susceptible to wilt. Wye Target,<br />
however, was wilt tolerant, immune to powdery mildew, <strong>and</strong> contained up to 12% -acid.<br />
In the early 1990s it accounted for almost 50% of the English hop acreage. Wye Viking<br />
<strong>and</strong> Wye Saxon were not successful; Yeoman was popular at first but had declined by<br />
1998 as had Zenith <strong>and</strong> Omega. By the end of the 20th century most hop-growing<br />
countries were breeding `super-alpha' hops with more than 15% -acid. Wye's<br />
contribution was Phoenix <strong>and</strong> Admiral, released in 1995. `Super-alpha' hops were wanted<br />
because it is more economic to process them than those with less -acid <strong>and</strong> 60% of the<br />
world hop crop is processed today. However, unless the world beer market increases<br />
dramatically, there will be no increased dem<strong>and</strong> for -acid so the growth of high-alpha<br />
hops will reduce the amount <strong>and</strong> area of hops required further.<br />
Most commercial hops are grown on wirework up to 7 m (23 ft.) high which is<br />
expensive to set up <strong>and</strong> maintain. Towards the end of the 20th century there was<br />
considerable interest in growing hops on a low trellis, not more than 3 m (10 ft.) high. The<br />
hops grow up plastic netting to produce a hedge which, besides being cheaper to set up,<br />
simplifies cultivation, spraying, <strong>and</strong> harvesting. Machines have been developed which<br />
straddle the rows <strong>and</strong>, in particular, pesticides can be applied in a much more controlled<br />
manner resulting in savings <strong>and</strong> much less environmental damage. In addition, crop<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
otation is easier with the cheaper support systems. Conventional varieties can be grown<br />
on these low trellises but special dwarf varieties, bred at Wye, are more successful. Dwarf<br />
hops contain a special gene which produces short internodal distances so the plants are<br />
half the height of conventional plants. The dwarf varieties released so far include First<br />
Gold (an aroma variety), Herald (high-alpha), Pioneer (a semi-dwarf dual purpose hop)<br />
<strong>and</strong> Pilot. The yields/hectare from these low trellises are comparable with those from<br />
st<strong>and</strong>ard wirework.<br />
In Germany traditional varieties were grown in five areas: Hallertau (85% of the<br />
German hop area), Spalt, Hersbruck, Tettnang <strong>and</strong> Jura. Since 1992 Jura has been<br />
incorporated into Hallertau <strong>and</strong> Elbe-Saale is a collective name given to the hop-growing<br />
areas of the former German Democratic Republic (Barth, 1999). Later, in addition to the<br />
traditional varieties, foreign varieties, Brewer's Gold, Northern Brewer <strong>and</strong> Record (from<br />
Belgium), with higher levels of -acid were also grown. Breeding was originally started<br />
at HuÈll, in 1926, to produce varieties resistant to downy mildew which retained the<br />
traditional aromas. The outbreak of wilt in the 1950s destroyed much of the major variety<br />
Hallertau mittelfruÈh but Northern Brewer <strong>and</strong> Hersbrucker spaÈt showed resistance to the<br />
German strain of Verticillium. Breeding continued to produce three new varieties: HuÈller<br />
Bitterer (commonly known as HuÈller), Hallertauer Gold, <strong>and</strong> Perle. HuÈller <strong>and</strong> Perle were<br />
resistant to both downy mildew <strong>and</strong> wilt <strong>and</strong> had higher levels of -acid than the<br />
traditional varieties. Hallertauer Gold was not resistant to wilt but both it <strong>and</strong> Perle were<br />
judged to have the same number of aroma fineness points as the traditional varieties.<br />
Taurus <strong>and</strong> Magnum are the latest German super-alpha varieties. The Saaz (Zatec) hop<br />
grown in Czechoslovakia is renowned for its fine aroma. The low yields it produces have<br />
been improved by clonal selection rather than breeding. The Saaz hop is thought to be<br />
related to the German varieties Tettnang <strong>and</strong> Spalter <strong>and</strong> the Japanese Shinshuwase. Hops<br />
were imported from Engl<strong>and</strong> to what was Yugoslavia <strong>and</strong> the Savinga (Styrian) Golding,<br />
imported by some brewers into Engl<strong>and</strong>, is identical with a seedless Fuggle. Yugoslavian<br />
hop breeding produced the `Super Styrian'; the high-alpha hops Atlas, Apolon, Ahil, <strong>and</strong><br />
Aurora. In Slovenia Blisk, Bobek <strong>and</strong> Buket have been produced <strong>and</strong> in the Backa region<br />
three more high-alpha hops were raised from Northern Brewer: Neoplanta, Vojvodina<br />
<strong>and</strong> Dunav.<br />
In the United States also the varieties grown have changed in the last twenty years.<br />
Eighty per cent of the US hops are grown in the Yakima valley Washington State, where<br />
Clusters (Early <strong>and</strong> Late) were the main variety <strong>and</strong> the st<strong>and</strong>ard kettle hop. The English<br />
varieties Fuggle, Bullion, <strong>and</strong> Brewers' Gold were also grown; the last two of these<br />
varieties were sometimes classified together as `English'. Clonal selections were made<br />
from Clusters <strong>and</strong> Talisman was released in 1968. Cascade, the first aroma hop from the<br />
US breeding programme, was released in 1972 <strong>and</strong> Willamette <strong>and</strong> Columbia in 1976.<br />
These last are triploid seedless aroma hops <strong>and</strong> Willamette still accounted for 14% of the<br />
US hop acreage in 2002. In the same year the high-alpha varieties Galena, Nugget,<br />
Columbus-Tomahawk, Zeus, Millenium <strong>and</strong> YCR-5 (Warrior) made up over half of the<br />
American hop area. Clusters accounted for only 3% but Eroica, Olympic <strong>and</strong> the English<br />
varieties had almost disappeared.<br />
Hops are grown in Australia in Victoria <strong>and</strong> Tasmania <strong>and</strong> in New Zeal<strong>and</strong> around<br />
Nelson on South Isl<strong>and</strong>. Being geographically isolated the hops produced in these<br />
countries are free of most of the pests <strong>and</strong> diseases found in the Northern hemisphere.<br />
Only the two-spotted mite (Tetranychus urticae Koch) <strong>and</strong> the red spider mite<br />
(Panonychus ulmi) occasionally give trouble. Thus, these countries can produce `organic'<br />
hops, without the use of pesticides, more easily than anywhere else in the world. Such<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
hops from New Zeal<strong>and</strong> are termed `Bio-Gro'. The major variety grown in Australia was<br />
Pride of Ringwood, bred by Nash from Salmon's Pride of Kent, which at one time<br />
occupied 90% of the Australian hop acreage. It is now being replaced by atriploid Super<br />
Pride <strong>and</strong> two super-alpha hops Opal <strong>and</strong> Victoria. Aroma hops occupy only 4% of the<br />
Australian hop area. In contrast, they occupy almost 50% of the New Zeal<strong>and</strong> acreage<br />
(Inglis,1999).NearlyallthehopsgrowninNewZeal<strong>and</strong>aretriploids,includingthedualpurpose<br />
hops Green Bullet, Sticklebract, Super Alpha, Southern Cross <strong>and</strong> Pacific Gem.<br />
The European aroma hops do not grow well in New Zeal<strong>and</strong> but from Hallertau<br />
mittelfruÈhtwo new aroma varieties more suited to New Zeal<strong>and</strong> conditions have been<br />
bred, NZ Hallertau Aroma <strong>and</strong> Pacific Hallertau. In Japan the traditional variety<br />
Shinshuwase was replaced by Kirin No. 2which, in turn, is being replaced by the superalpha<br />
varieties Toyomidori, Kitamidori <strong>and</strong> Eastern Gold.<br />
The characteristics of the main varieties are collected in Table 7.3. These data were<br />
collected from many sources <strong>and</strong> may not be strictly comparable. No doubt in twenty<br />
years time different varieties will be grown. Plant breeders continue to seek new varieties<br />
with increased resin content, increased disease resistance <strong>and</strong> better yields. Other goals<br />
relate to resin quality. Some brewers think that humulone gives a better bitter flavour than<br />
cohumulone so require hops in which the proportion of cohumulone in the a-acids is as<br />
low as possible. Some cultivars deteriorate on storage more rapidly than others <strong>and</strong>, since<br />
all hops cannot be processed immediately after harvest, good storage stability is<br />
desirable. Although many new varieties show resistance to fungal diseases most growers<br />
still have to use pesticides to control aphids <strong>and</strong> mites. New breeding programmes have<br />
produced aphid-resistant hops that can be grown `organically'. The hop is a long-day<br />
plant which grows between 30 <strong>and</strong> 55ë of latitude; when grown nearer the equator<br />
artificial illumination is necessary. Countries which grow hops under these conditions are<br />
trying to breed varieties adapted to the shorter day length.<br />
The unambiguous characterization of hop cultivars is difficult although methods based<br />
on morphology <strong>and</strong> chemical analysis usually give good indications. However, methods<br />
based on DNA analysis reflect the genotype of the cultivar irrespective of the stage of<br />
plant development, the environmental or disease status. Briefly a sample of hop DNA is<br />
subjected to a r<strong>and</strong>om amplified polymorphic DNA (RAPD) process using arbitrary 10<br />
mer primers. The products are then separated by electrophoresis followed by staining,<br />
ultraviolet visualization <strong>and</strong> DNA sequencing. Using this method Murakami (2000)<br />
produced a dendrogram, based on genetic distance, which resolved most of the common<br />
varieties into six clusters; most of the high-alpha hops were in the first cluster.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 7.3 Properties of principal hop cultivars a<br />
Cultivar -acid (%) = Cohumulone Oil (%) Humulene/ Comments<br />
ratio (%) Caryophyllene<br />
ratio<br />
Australia<br />
Opal 13.0 3.2 30 1.5 2.5<br />
Pride of Ringwood b<br />
9.0±11.0 1.7 33 2.0 0.1<br />
Super Pride 13.9 2.4 27 1.0 0.4<br />
Topaz 11.5 2.0 40 1.0 0.2<br />
Victoria 11.0±14.0 2.1 38 1.1 1.6<br />
Belgium<br />
Record 5.5±8.5 1.0 30 1.8 2.5<br />
Czech Republic<br />
Bor 6.5±11 1.8 25 1.5 3.2 Dual<br />
Premiant 7.0±11.0 1.9 22 1.5 3.0 Dual<br />
Saaz 3.0±4.5 0.9 26 0.4 3.5<br />
Sladek 4.0±8.0 1.0 28 1.5 2.4 Aroma<br />
Engl<strong>and</strong> c<br />
Admiral 13.5±16.2 2.6±3.2 26±32 1.0±1.7 Wilt tolerant<br />
Bramling Cross 6.0±7.8 2.4±3.1 26±31 0.7±1.0 2.2 Wilt tolerant<br />
Brewers Gold 5.5±8.5 1.9 38 1.5 2.3<br />
Bullion 6.0±9.0 1.9 36 3.2 1.5<br />
First Gold (Dwarf) 5.6±8.7 2.4±3.2 29±34 0.7±1.4 3.2 Wilt tolerant<br />
Fuggle 3.0±5.6 1.5±2.2 29±30 0.7±1.1 3.3<br />
Goldings 4.4±6.7 2.1±2.6 26±32 0.8±1.0 3.5<br />
Herald (Dwarf) 11.0±13.0 2.4 37 1.0±2.2 2.4 Wilt resistant<br />
Northern Brewer 6.5±10.0 2.0 23 2.0 2.8<br />
Phoenix 12.0±15.0 2.1±2.6 24±28 1.2±2.5 Wilt resistant<br />
Progress 6.0±7.5 2.8±3.3 27±36 0.5±0.8 3.3 Wilt tolerant<br />
Whitbread Golding<br />
Variety (WGV) 5.4±7.7 2.3±3.0 32±43 0.8±1.2 3.5 Wilt tolerant<br />
Wye Challenger 6.5±8.5 1.8±2.1 20±25 1.0±1.5 3.1<br />
Wye Northdown 6.8±9.6 1.5±2.2 24±29 1.2±2.2 2.7<br />
Wye Target 9.9±12.6 2.2±2.8 35±39 1.2±1.4 2.4 Wilt resistant<br />
France<br />
Strisselspalt 2.0±5.0 1.0 24 0.7 2.4<br />
Germany d<br />
Hallertau mittelfruÈh 3.5±5.0 1.0 21 0.6±1.0 3.7 Aroma<br />
Hallertau tradition 5.0±7.0 1.2 26 1.2±1.4 3.8 Aroma<br />
Hersbrucker spaÈt 3.0±5.0 0.9 25 0.6±1.0 2.5 Aroma<br />
HuÈller (Bitterer) 5.0±7.0 1.2 30 1.0±1.3 1.9<br />
Magnum 12.0±14.0 2.6 26 1.6±2.1 3.6 Bitter<br />
Perle 6.0±8.5 1.6 29 1.0±1.3 2.6 Aroma<br />
Spalter 4.0±5.0 1.1 24 0.6±1.0 3.3<br />
Spalter 4.0±5.5 1.3 24 0.6±1.0 2.0 Aroma<br />
Taurus 12.0±15.0 3.0 24 1.2±1.5 3.7 Bitter<br />
Tettnang 3.5±5.0 1.0 25 0.4±1.0 2.7 Aroma<br />
New Zeal<strong>and</strong> e<br />
Green Bullet 12.5±13.5 1.8 42 0.8 3.2<br />
NZ Hallertau aroma 8.5 1.3 29 1.25 2.0<br />
Pacific Gem 14.0±16.0 1.8 41 1.5 3.2<br />
Pacific Hallertau 6.0 1.0 26 1.26 2.9<br />
Southern Cross 11.0±12.0 1.9 27 1.2 3.4<br />
Stricklebract 13.5±14.5 1.7 38 1.0 3.2<br />
Super Alpha 12.5±13.5 1.5 38 1.5 3.2<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 7.3 (Continued).<br />
Cultivar -acid (%) = Cohumulone Oil (%) Humulene/ Comments<br />
ratio (%) Caryophyllene<br />
ratio<br />
Pol<strong>and</strong> f<br />
Limbus 5.3 ± 36 1.7 2.2 Aroma<br />
Lubekski 4.0 ± 31.8 1.2 5.1 Aroma<br />
Lublin 3.5±4.5 1.3 27 1.0 3.7<br />
Marynka 11.1 ± 25 2.3 5.0 Bitter<br />
Oktawia 10.6 ± 34 1.6 4.4 Bitter<br />
Sybilla 7.3 ± 34 1.7 3.8 Bitter<br />
Zbyszko 8.5 ± 26 1.0 2.5 Bitter<br />
Slovenia<br />
Ahil 8.0±10.0 2.2 27 1.0 2.4<br />
Apolon 8.0±10.0 2.2 27 1.0 2.5<br />
Atlas 8.0±10.0 2.2 31 0.8 2.3<br />
Aurora 8.5±10.5 2.1 25 1.0 2.9<br />
Bobek 5.7 1.1 33 2.3 3.2<br />
Blisk 6.0 2.0 37 1.9 2.4<br />
Buket 8.2 2.3 24 2.7 2.9<br />
Cekin 5.6 2.2 27 1.3 3.3<br />
Celeia 5.4 2.1 28 2.1 2.4<br />
Cerera 5.2 1.7 30 2.0 2.7<br />
Cicero 8.7 2.9 28 1.5 2.9<br />
Savinja (Styrian)<br />
Golding 4.5±6.0 2.0 28 0.8 3.1<br />
South Africa g<br />
Outeniqua 12.0±13.5 2.8 29 1.6 3.0 Bitter<br />
Southern Brewer 9.0±10.5 2.4 39 1.5 2.1 Bitter<br />
Southern Promise 9.5±11.5 2.3 21 0.7 2.4 Dual purpose<br />
Southern Star 12.0±15.5 2.8 31 1.6 1.5 Bitter<br />
United States h<br />
Cascade 4.5±7.0 1.0 37 1.2 2.7<br />
Chelan 14.5 1.4 35 1.5 1.2<br />
Chinook 12.0±14.0 3.9 32 2.0 2.<br />
Cluster 5.5±8.5 1.4 36±42 0.4±0.8 2.5<br />
Columbus/Tomohawk/<br />
Zeus (CTZ) 14±18 3.1 29±34 2.0±3.5 1.7<br />
Galena 11.0±13.0 1.6 44 0.9±1.4 2.0<br />
Horizon 13.6 2.2 19 1.9 1.6 Dual<br />
Millenium 15.5 3.2 30 2.0 2.4<br />
Mount Hood 5.0±8.0 1.1 23 1.1 2.5<br />
Nugget 12.0±14.0 3.3 27 2.0 2.2<br />
Willamette 5.0±7.0 1.6 33 1.2 2.9<br />
YCR-5 (Warrior) 14.5±16.5 2.6 24±26 1.0±2.0 1.9<br />
Humulus lupulus var.<br />
neomexicanus 3.1 1.4 65 0.6±0.8 0.73<br />
a<br />
From Neve (1991) <strong>and</strong> Darby (personal communication). Neve also gives data for Omega <strong>and</strong> Yeoman (GB),<br />
Orion (D) <strong>and</strong> Aquila, Banner, Olympic <strong>and</strong> Talisman (USA).<br />
b<br />
Leggett (2004).<br />
c<br />
National Hop Association of Engl<strong>and</strong> (2000).<br />
d<br />
Barth (1999).<br />
e<br />
Inglis (1999).<br />
f<br />
Brudzynski <strong>and</strong> Baranowski (2003).<br />
g<br />
Brits <strong>and</strong> Linsley-Noakes (2001).<br />
h<br />
www.usahops<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
7.8 References<br />
BARTH, S. J. (1999) Ferment, 12 (5), 40.<br />
BARTH, H. J., KLINKE, C. <strong>and</strong> SCHMIDT, C. (1994) The Hop Atlas ± The History <strong>and</strong> Geography of the<br />
Cultivated Plant. Barth, NuÈrnberg, 383 pp.<br />
BEHRE, K.-E. (1999) Vegetation History <strong>and</strong> Archaeobotany, 8, 35±48.<br />
BENITEZ, J. L., FORSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. <strong>and</strong> WESTWOOD,<br />
K. T. (1997). European Brewery Convention ± Manual of Good Practice: Hops <strong>and</strong> Hop Products,<br />
pp. xiv + 186. Verlag Hans Carl, NuÈrnberg.<br />
BRITS, G. <strong>and</strong> LINSLEY-NOAKES, G. C. (2001) Proc. 8th. Conv. Inst. <strong>Brewing</strong>, Africa Section, Sun City,<br />
South Africa, p. 176.<br />
BRUDZYNSKI, A. <strong>and</strong> BARANOWSKI, K. (2003) J. Inst. <strong>Brewing</strong>, 109, 154.<br />
BURGESS, A. H. (1964) Hops: Botany, Cultivation <strong>and</strong> Utilization. Leonard Hill, London, pp. xx + 300.<br />
DE KEUKELEIRE, J., OMMS, G., HEYERICK, A., ROLDAN-RUIZ, I., VAN BOCKSTAELE, E. <strong>and</strong> DE KEUKELEIRE,<br />
D. (2003) J. Agric. Food Chem., 51, 4436.<br />
EUROPEAN BREWERY CONVENTION (1987) Monograph XIII. Symposium on Hops, Weihenstephan, pp. xii<br />
+ 286.<br />
EUROPEAN BREWERY CONVENTION (1994). Monograph XXI. Symposium on Hops, Zoeterwoude,<br />
pp. xviii + 300.<br />
EUROPEAN BREWERY CONVENTION (1997) see Benitez et al.<br />
FILMER, R. (1982) Hops <strong>and</strong> Hop Picking, Shire Publications, Princes Risborough, 80 pp.<br />
GREEN, C. P. (1978) J. Inst. <strong>Brewing</strong>, 84, 312.<br />
INGLIS, T. (1999). Ferment, 12 (5), 19±28.<br />
JOHNSTONE, D. I. H. (1997) Ferment, 10, 325±329.<br />
LEGGETT, G. (2004) Brewer Intern., 4 (2), 42.<br />
LEMUSIEAU, G., LIEÂ GEOIS, C. <strong>and</strong> COLLIN, S. (2001) Food Chemistry, 72, 413.<br />
LIKENS, S. T., NICKERSON, G. B., HAVOULD, A. <strong>and</strong> ZIMMERMAN, C. E. (1978) Crop Sciences, 18, 380±386.<br />
MOIR, M (2000) J. Amer. Soc. Brew. Chem., 58, 131±146.<br />
MURAKAMI, A. (2000) J. Inst. <strong>Brewing</strong>, 106, 157±161.<br />
NATIONAL HOP ASSOCIATION OF ENGLAND (2000). The Hop Guide, 22 pp.<br />
NEVE, R. A. (1991) Hops. Chapman & Hall, London, xii + 266 pp.<br />
PICKETT, J. A., COATES, J. <strong>and</strong> SHARPE, F. R. (1977) Proc. 16th Congr. Eur. Brew. Conv., Amsterdam,<br />
p. 123; J. Inst. <strong>Brewing</strong>, 83, 302.<br />
WILSON, D. G. (1975) New Phytologist, 75, 627±648.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
8<br />
The chemistry of hop constituents<br />
8.1 Introduction<br />
Freshly picked hop cones contain about 80% moisture <strong>and</strong> rapidly go mouldy if not dried<br />
on a kiln or oast. Commercial hops contain c. 10% moisture. Moisture in hops is usually<br />
measured as the loss on drying at 105±107 ëC for 1 hour (Analytica-EBC, 1998). For green<br />
hops a longer period of drying (four hours) is necessary. Some essential oil may be lost<br />
during oven drying. Alternative methods for estimating moisture in hops include drying in a<br />
vacuum desiccator or azeotropic distillation (Dean <strong>and</strong> Stark method). Hop analyses are<br />
usually reported `as is' but may occasionally be given with reference to dry matter.<br />
Most of the brewing value of the hop is found in the resins <strong>and</strong> essential oils which are<br />
only slightly soluble in water. However, Goldstein et al. (1999) showed that hops contain<br />
20±25% of water-soluble constituents which dissolve directly in the boiling wort. This<br />
fraction will include carbohydrates, amino acids, proteins, polyphenols, <strong>and</strong> inorganic<br />
salts. MacWilliam (1953) showed that hops contain c. 2% of sugars mainly fructose,<br />
glucose <strong>and</strong> raffinose; hops also contain 1±2% of pectin. Goldstein et al. (1999) drew<br />
attention to the presence of glycosides in the water-soluble fraction of hops. Most<br />
organisms transport water-insoluble substances by conjugating them with a sugar, usually<br />
glucose, to produce a water-soluble glucoside. -Sitosteryl glucoside was found in hops<br />
as long ago as 1913 (Power et al., 1913) <strong>and</strong> many polyphenols are found as glycosides in<br />
hops (see later). Goldstein et al. (1999) also found that many volatile constituents of hops<br />
are also present in the water-soluble fraction bound as glycosides.<br />
Hops contain 2.0±3.5% of nitrogen equivalent to 12.5±21.7% of protein. About 0.5%<br />
of the nitrogen, equivalent to 3.1% of protein, is soluble in water. Hops contain c. 0.1% of<br />
amino acids <strong>and</strong> dried hops yield about 8% of ash (inorganic matter). Lipids ± oils, fats,<br />
<strong>and</strong> waxes ± are, like the hop resins, insoluble in water. Hop seeds contain up to 32% of<br />
triglycerides but they are not usually dispersed from intact seeds during wort boiling. Hop<br />
wax is derived from the cuticle of cones <strong>and</strong> leaves <strong>and</strong> is a mixture of long chain<br />
hydrocarbons (C29 predominates), alcohols, acids <strong>and</strong> esters together with -sitosterol.<br />
Although these materials are only slightly soluble in water they may be included with the<br />
resins in solvent extracts of hops.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The chemistry of hop constituents was reviewed by Stevens (1967) <strong>and</strong> Moir (2000)<br />
provided amillenium review. Other reviews have concentrated on individual classes of<br />
hop compound.<br />
8.2 Hop resins<br />
8.2.1 Introduction<br />
A book, Chemistry <strong>and</strong> Analysis of Hop <strong>and</strong> Beer Bitter Acids (Verzele <strong>and</strong> De<br />
Keukeleire, 1991) provides adetaileddiscussionofmostofthereactions discussedinthis<br />
section. Later information is found in the European Brewery Convention Symposium on<br />
Hops, Zoeterwoude (1994) <strong>and</strong> the European Brewery Convention: Manual of Good<br />
Practice-Hops <strong>and</strong> Hop Products (Benitez et al., 1997). The Nomenclature Committee of<br />
the Hops Liaison Committee (1969) made recommendations defining A, non-specific<br />
fractions <strong>and</strong> B, specific compounds <strong>and</strong> mixtures of specific compounds.<br />
A. Non-specific fractions<br />
1. Total resins The part of the hop constituents that is characterized by solubility both<br />
in cold methanol <strong>and</strong> diethyl ether (mainly hard resins, uncharacterized soft resins,<br />
-acids <strong>and</strong> -acids). The requirement that the total resins should be soluble in cold<br />
methanol is designed to exclude hop wax which will slowly crystallize from cold<br />
methanol.<br />
2. Total softresins The fraction ofthetotal resins thatischaracterizedbysolubilityin<br />
hexane (mainly -acids, -acids <strong>and</strong> uncharacterized soft resins).<br />
3. Hard resins The fraction of the total resins that is characterized by insolubility in<br />
hexane. It is calculated as the difference between total resins <strong>and</strong> total soft resins.<br />
4. -Fraction The total soft resins minus the -acids.<br />
5. Uncharacterized soft resins That portion of the total soft resins that has not been<br />
characterized as specific compounds.<br />
B. Specific compounds <strong>and</strong> mixtures of specific compounds<br />
1. The -acids (8.1). These are mainly humulone (8.1a), cohumulone (8.1b) <strong>and</strong><br />
adhumulone (8.1c).<br />
2. The -acids (8.2). These are mainly lupulone (8.2a), colupulone (8.2b) <strong>and</strong><br />
adlupulone (8.2c).<br />
3±8.Chemical descriptions of the -<strong>and</strong> -acid analogues are given in Table 8.1.<br />
9. The iso- -acids (8.40). These are mainly isohumulone, isocohumulone <strong>and</strong> isoadhumulone.<br />
10. Isohumulone (8.40a). The mixture of cis- <strong>and</strong> trans-isohumulone. Similarly,<br />
isocohumulone (8.40b) refers to a mixture of cis- <strong>and</strong> trans-isocohumulone <strong>and</strong><br />
isoadhumulone (8.40c) to a mixture of cis- <strong>and</strong> trans-isoadhumulone,<br />
11. cis-Isohumulone (8.43). The iso- -acid with the empirical formula C21H30O5. It is<br />
an oil with the higher partition coefficient in a phase system of a hydrocarbon <strong>and</strong> a<br />
buffer, <strong>and</strong> contains an isovaleryl side chain. Cis- means that the 3-methyl-2-butenyl<br />
side chain <strong>and</strong> the tertiary hydroxyl group are on the same side of the ring.<br />
12. trans-Isohumulone (8.44). The iso- -acid with the empirical formula C21H30O5,<br />
with a m.p. 72 ëC <strong>and</strong> the lower partition coefficient in a phase system of a<br />
hydrocarbon <strong>and</strong> a buffer, <strong>and</strong> contains an isovaleryl side chain. Trans-means that<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the 3-methyl-2-butenyl side chain <strong>and</strong> the tertiary hydroxyl group are on opposite<br />
sides of the ring.<br />
13. cis-Isocohumulone. As in 11 but with reference to C20H28O5 <strong>and</strong> an isobutyryl<br />
(R ˆ Pr i ) side chain.<br />
14. trans-Isocohumulone. As in 12 but with reference to C 20H 28O 5 <strong>and</strong> an isobutyryl<br />
side chain.<br />
15. cis-Isoadhumulone. As in 11 but with reference to a 2-methylbutyryl<br />
(R ˆ CHMeEt) side chain.<br />
16. trans-Isoadhumulone. As in 12 but with reference to a 2-methylbutyryl side chain.<br />
17. Allo-Iso- -acids (8.41). These are isomers of the iso- -acids having a shifted<br />
double bond in the isohexenoyl side chain (i.e. 4-methyl-2-pentenoyl). Of each alloiso-<br />
-acid there is a cis- <strong>and</strong> a trans- form. The following specific names are<br />
therefore proposed: cis-allo-isohumulone, trans-allo-isohumulone, cis-allo-isocohumulone,<br />
trans-allo-isocohumulone, cis-alloisoadhumulone <strong>and</strong> trans-allo-isoadhumulone.<br />
18. Hulupones (8.85). These consist of hulupone, cohulupone <strong>and</strong> adhulupone.<br />
19. Hulupone (8.85a). Has the empirical formula C20H28O4. It is 2,2-di[3-methyl-2butenyl]-5-isovaleryl-1,2,4-cyclopentanetrione<br />
<strong>and</strong> is formed from lupulone.<br />
20. Cohulupone (8.85b). As in 19 but with reference to C19H26O4 <strong>and</strong> a 5-isobutyryl<br />
side chain.<br />
21. Adhulupone (8.85c). As in 19 but with reference to C 20H 28O 4 <strong>and</strong> a 5-[2methylbutyryl]<br />
side chain.<br />
22. Humulinic acids (8.3). These consist of the cis-<strong>and</strong> trans- forms of humulinic acid,<br />
cohumulinic acid <strong>and</strong> adhumulinic acid.<br />
23. cis-Humulinic acid. Has the empirical formula C15H22O4 with m.p. 68 ëC <strong>and</strong> the<br />
higher partition coefficient in a phase system of a hydrocarbon <strong>and</strong> a buffer. Cismeans<br />
that the 3-methyl-2-butenyl side chain <strong>and</strong> the alcoholic hydroxyl are on the<br />
same side of the ring.<br />
24. trans-Humulinic acid. Has the empirical formula C 15H 22O 4 with m.p. 95 ëC <strong>and</strong> the<br />
lower partition coefficient in a phase system of a hydrocarbon <strong>and</strong> a buffer. Transmeans<br />
that the 3-methyl-2-butenyl side chain <strong>and</strong> the alcoholic ring hydroxyl group<br />
are on opposite sides of the ring. Similar considerations will apply to ciscohumulinic<br />
acid, trans-cohumulinic acid, cis-adhumulinic acid <strong>and</strong> trans-adhumulinic<br />
acid.<br />
The sub-committee also recommended that where the intention is to refer to `beer bitter<br />
substances', usually an incompletely known mixture, this phrase should be used. Using<br />
these definitions the bulk of the brewing <strong>and</strong> bittering value of the hop is found in the<br />
total soft resins <strong>and</strong>, in particular, in the -acids. Only traces of the -acids survive into<br />
beer; they are transformed during wort boiling into the iso- -acids which are the major<br />
bittering principles of beer. The importance of the allo-iso- -acids is still debatable. The<br />
-acids are too insoluble in water to contribute to beer flavour themselves but they can be<br />
oxidized into hulupones which are bitter <strong>and</strong> are minor bittering principles in some beers.<br />
Hydrolysis of the -acids <strong>and</strong> the iso- -acids gives a mixture of humulinic acids which<br />
are not bitter. The humulinic acids are not normally found in beer but they may be present<br />
in some isomerized hop extracts <strong>and</strong> beers brewed therefrom.<br />
The -acids can be separated from the total soft resins by their ability to form an<br />
insoluble lead salt (chelate?) with lead(II) acetate in methanol; the -fraction may be<br />
recovered from the mother liquors. The -acids can be regenerated from the lead salts by<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
suspending them in methanol <strong>and</strong> adding either sulphuric acid or hydrogen sulphide gas.<br />
After removal of the inorganic matter by filtration, evaporation of the solvent leaves a<br />
mixture of -acids from which humulone may slowly crystallize. Humulone (8.1a) was<br />
the only -acid known until 1953 when Rigby <strong>and</strong> Bethune isolated the analogues<br />
cohumulone(8.1b)<strong>and</strong>adhumulone(8.1c)bycountercurrentdistribution.Traces ofother<br />
analogues have been found (Table 8.1). Later, Hermans-Lockkerbol <strong>and</strong> Verpoorte<br />
(1994) used centrifugal partition chromatography to separate the -acids. Most hop<br />
varieties contain about 10% of adhumulone in their -acids but the proportion of<br />
cohumulone appears to be a varietal characteristic (Table 7.3). The -acids form<br />
crystalline 1:1 complexes with 1,2-diaminobenzene (o-phenylenediamine). Repeated<br />
recrystallization concentrates the humulone complex with respect to those of the other -<br />
acids. Decomposition of the yellow complex with 2N-hydrochloric acid followed by<br />
recrystallization from cyclohexane at 20ëC gives humulone, m.p. 63ëC. Most of the<br />
chemistry of the -acids has been carried out on humulone purified in this way but<br />
Simpson (1993a) found that asample prepared in this manner still contained 8% of<br />
cohumulone <strong>and</strong> 1% of adhumulone.<br />
The -acids (8.2) may crystallize from the -fraction obtained after the lead salts have<br />
been precipitated. In <strong>practice</strong> it may be better to dilute the methanol solution with brine<br />
<strong>and</strong> extract the -fraction into light petroleum. Alternatively, asolution of ahop extract<br />
inhexanemaybeextractedfirstwithdisodiumcarbonate,toremovethestronger -acids,<br />
<strong>and</strong> then with sodium hydroxide to recover the -acids. From the mixture of -acids<br />
Lermerisolatedlupulone(8.2a)in1863.Muchlater,inthe1950s,itwasfoundthatthe -<br />
acidwhichcrystallizedfromEnglishhopextractswascolupulone(8.2b)Thus,likethe -<br />
acids, the -acids are amixture of analogues. They are too sensitive to aerial oxidation to<br />
be separated by countercurrent distribution. They can be separated by HPLC, but before<br />
this technique was available, the proportions of the -acid analogues was found by<br />
converting them to tetrahydro- -acids (8.9, Fig 8.1) which were stable during<br />
countercurrent distribution. It was found that the -acids were always richer in<br />
colupulone than the -acids were in cohumulone. Indeed aregression equation was<br />
obtained:<br />
%Colupulone in -acids =0.943 (% cohumulone in -acids) +20.2<br />
The structures of humulone (8.1a) <strong>and</strong> lupulone (8.2a), except in minor detail, were<br />
worked out by Wollmer <strong>and</strong> Wiel<strong>and</strong> (Fig. 8.1). Both are acylphloroglucinols substituted<br />
with 3-methyl-2-butenyl- (dimethylallyl- or isoprenyl-) groups, three in lupulone <strong>and</strong> two<br />
in humulone, which also has a tertiary hydroxyl group. Hydrolysis of humulone,<br />
C21H30O5, gives humulinic acid (8.3a), C15H22O4, <strong>and</strong> 4-methyl-3-pentenoic acid<br />
(isohexenoic acid, (8.13)), C6H10O2, which accounts for all the carbon atoms, but<br />
isobutyraldehyde (8.14) is also formed. Hydrogenation of humulinic acid gave<br />
dihydrohumulinic acid (8.4a) which by Clemmersen reduction gave 1, 3-di-isopentylcyclopentane<br />
(8.11) establishing the five-membered ring in humulinic acid. Mild<br />
hydrogenation of humulone gave the tetrahydro-derivative (8.13a), showing two double<br />
bonds, but with palladium chloride hydrogenolysis occurs giving humuloquinol (8.5a),<br />
C16H24O5, which is readily oxidized to humuloquinone (8.6a). Hydrolysis of<br />
humuloquinone gives isohumulinic acid (8.7a) also obtained from dihydrohumlinic acid<br />
(8.4a) by oxidation with bismuth oxide. Similarly, mild hydrogenation of lupulone (8.2a)<br />
gives a hexahydro-derivative but hydrogenolysis gives tetrahydrodeoxyhumulone (8.8a)<br />
which, when shaken in air or oxygen with lead acetate solution, gives the lead salt of<br />
tetrahydrohumulone (8.9a) linking the - <strong>and</strong> -acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 8.1 Analogues of the - <strong>and</strong> -acids<br />
-acids -acids<br />
Acyl side chain (R) Name Formula m.p. (ëC) [ ] 24<br />
0 pKa Name Formula m.p. (ëC)<br />
a -CO.CH 2.CH(CH 3) 2 Humulone C 21H 30O 5 64.5ë 211ë 5.5 Lupulone C 26H 38O 4 92ë<br />
isovaleryl<br />
b -CO.CH(CH 3) 2 Cohumulone C 20H 28O 5 oil 208.5ë 4.7 Colupulone C 25H 36O 4 93 94ë<br />
isobutyryl<br />
c -CO.CH(CH 3).CH 2.CH 3 Adhumulone C 21H 28O 5 oil 187ë 5.7 Adlupulone C 26H 38O 4 82 83ë<br />
2-methylbutyryl<br />
d -CO.CH2.CH3 Posthumulone a<br />
propionyl<br />
e -CO.CH2.CH2.CH(CH3)2 Prehumulone b<br />
4-methylpentanoyl<br />
f -CO.(CH2)4.CH3 Adprehumulone c<br />
hexanoyl<br />
C19H26O5 oil ± ± ± d<br />
C22H32O5 oil 172ë ± ± e<br />
C22H32O5 ± ± ± ± e<br />
C24H34O4<br />
C27H40O4<br />
C27H40O4<br />
g -CO.CH2.CH2.CH(CH3).CH2.CH3 ± C23H34O5 ± ± ± ± C28H42O4 91ë<br />
4-methylhexanoyl<br />
a Verzele (1958).<br />
b Rillaers <strong>and</strong> Verzele (1962).<br />
c Smith et al. (1998).<br />
d Riedl et al. (1956).<br />
e Riedl (1954).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
101ë<br />
91ë<br />
90ë
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Fig. 8.1 Reactions of - <strong>and</strong> -acids.
In both humulone (8.1a) <strong>and</strong> lupulone (8.2a) carbon atom-6 in the ring is sp3<br />
hybridized; the other atoms in the ring are sp2 hybridized. In humulone carbon atom-6 is<br />
substituted by four different groups <strong>and</strong> so it is chiral <strong>and</strong> humulone is optically active.<br />
The natural -acids are laevorotatory <strong>and</strong> their specific angular rotations are given in<br />
Table 8.1. Using the sequence rules of Cahn et al. (1951, 1956 <strong>and</strong> 1966) the absolute<br />
stereochemistry of the -acids is R(ectus) so natural humulone is R-( )-humulone.<br />
Synthetic racemic humulone is an equimolecular mixture of R-( )-humulone <strong>and</strong><br />
unnatural S-(+)-humulone <strong>and</strong>, as such, will not rotate the plane of polarized light. Fairly<br />
mild chemical treatment will convert R-( )-humulone into racemic RS-( )-humulone.<br />
( )-Humulone has not been resolved into its enantiomers but this should be possible.<br />
In lupulone (8.2a) <strong>and</strong> the other -acids carbon atom-6 is substituted with two 3methyl-2-butenyl-<br />
(isoprenyl-) groups so it will not be chiral <strong>and</strong> will not rotate the plane<br />
of polarized light. However, in adlupulone (8.2c) there is achiral centre at C-2 of the 2methylbutyryl-side<br />
chain so natural adlupulone should be optically active. It follows that<br />
natural adhumulone (8.1c) has two chiral centres so four enantiomers are possible. By<br />
analogy with L-isoleucine, the 2-methylbutyryl-side chain in natural adhumulone <strong>and</strong><br />
adlupulone probably has the S-configuration.<br />
It is difficult to write a single structure for most of the hop resins because they exhibit<br />
keto-enol tautomerism where ketones exist in equilibrium with the related enol:<br />
With an isolated ketone, as in acetone (8.15), the equilibrium lies well on the ketone side:<br />
However, -dicarbonyl compounds such as acetylacetone (pentane-2, 4-dione, (8.16))<br />
exist largely in the enol form which is stabilized by hydrogen bonding.<br />
Even more tautomers can be drawn for -tricarbonyl compounds such as triacetylmethane<br />
(8.17)<br />
(CH 3.CO) 3.CH<br />
(8.17) triacetylmethane<br />
Most of the hop resins contain -di- or -tricarbonyl systems which are enolized. In most<br />
cases the individual tautomers cannot be isolated but it is usually possible to estimate the<br />
proportions of the individual tautomers by proton magnetic resonance spectroscopy<br />
(PMR) or other physical methods. Irrespective of the main component of a tautomeric<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mixture it can react both as a ketone <strong>and</strong> as an enol. It is an equilibrium mixture so that if<br />
one component is removed by reaction it will be regenerated according to the<br />
equilibrium. Thus, for example, phloroglucinol, the parent of the hop resins, exists almost<br />
entirely in the trienol form (8.18) <strong>and</strong> can form a triacetate (8.20) but it can react as<br />
cyclohexane-1, 3, 5-trione (8.19) to produce a tri-oxime (8.21).<br />
O·Ac<br />
AcO O·Ac<br />
(8.20)<br />
phloroglucinol<br />
triacetate<br />
OH<br />
HO OH<br />
(8.18)<br />
phloroglucinol<br />
O<br />
O O<br />
(8.19)<br />
cyclohexane-<br />
1,3,5-trione<br />
HO·N<br />
N·OH<br />
(8.21)<br />
cyclohexane-<br />
1,3,5-trioxime<br />
N·OH<br />
PMR measurements show that humulone exists principally as a dienol. Of the possible<br />
tautomeric structures (8.1.1±4), structure (8.1.1) is thought to represent the major<br />
tautomer; structures (8.1.2 ) <strong>and</strong> (8.1.4) were excluded on the basis of optical rotatory<br />
dispersion measurements <strong>and</strong> (8.1.1) was preferred over (8.1.3) by comparison with<br />
model compounds (De Keukeleire <strong>and</strong> Verzele, 1970).<br />
OH<br />
O<br />
OH OH<br />
R' CO·R R' CO·R R' CO·R R'<br />
HO<br />
O<br />
HO OH<br />
O<br />
OH<br />
O O<br />
HO R' HO R' HO R' HO R'<br />
(8.1.1) (8.1.2) (8.1.3) (8.1.4)<br />
Similarly, PMR measurements suggest that lupulone exists as a mixture of two tautomers<br />
(8.2.1) <strong>and</strong> (8.2.2) in the ratio 7:3 (Collins et al., 1971).<br />
HO<br />
H<br />
O O<br />
O<br />
R HO<br />
(8.2.1) (8.2.2)<br />
Consideration of the tautomers of acetylacetone (8.16) shows that the enol hydrogen atom<br />
is bound to different oxygen atoms in the two cyclic forms. This hydrogen can dissociate<br />
leaving an enolate ion stabilized by resonance. Thus, -di- <strong>and</strong> -tri-carbonyl compounds<br />
are acids <strong>and</strong> these functions provide the acidity of the hop resins. The strength of acids<br />
can be compared on the pKa scale where<br />
pKa= log 10Ka<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
R<br />
O<br />
OH<br />
C<br />
O<br />
H<br />
R
<strong>and</strong> Ka is the dissociation constant of the acid. On this scale completely dissociated<br />
mineral acids have negative values while carboxylic acids have values such as: methanoic<br />
(formic) acid, pKa 3.77, ethanoic (acetic) acid, pKa 4.76, propanoic acid, pKa 4.88, <strong>and</strong><br />
benzoic acid, pKa 4.20. Triacetylmethane (8.17), pKa 5.81, is a stronger acid than<br />
acetylacetone (8.16) pKa 8.13, while phenol, pKa 10.0, is weaker still. Phloroglucinol<br />
(8.18) has three phenolic hydroxy groups; the pKas of the first two are 7.97 <strong>and</strong> 9.23.<br />
pKa measurements should be made in dilute aqueous solution but, because of their<br />
limited solubility, early estimates of the pKas of the hop resins were made in aqueous<br />
methanol solutions giving pKa values for humulone, 5.5, cohumulone, 4.7, <strong>and</strong> adhumulone,<br />
5.7. In aqueous solutions Simpson <strong>and</strong> Smith (1992) found equilibrium pKa values for the<br />
most acidic functions were: humulone, 5.0, colupulone (8.2b), 6.1, trans-isohumulone<br />
(8.44), 3.1; <strong>and</strong> trans-humulinic acid (8.3a), 2.7. Thus, the - <strong>and</strong> -acids are weaker than<br />
carboxylic acids, such as acetic acid, but their isomerization products, the iso- -acids, are<br />
stronger. It should be recalled that when the pH of the medium equals the pKa of the acid<br />
50% of the acid will be present as the anion <strong>and</strong> 50% undissociated.<br />
The solubilities of humulone, lupulone <strong>and</strong> (trans-)humulinic acid were measured by<br />
Spetsig (1955) (Fig. 8.2). As expected, the solubilities increase with temperature <strong>and</strong> increasing<br />
pH; in each case the anion is more soluble than the undissociated acid. For example, in boiling<br />
wort at pH 5.0, about 200 mg/litre of humulone will dissolve but, if no transformation takes<br />
place, most of this will be precipitated from conditioned beer (pH 4.0) at 0 ëC.<br />
mg/l<br />
5000<br />
1000<br />
500<br />
100<br />
50<br />
10<br />
5<br />
1<br />
100°<br />
40°<br />
25°<br />
0°<br />
100°<br />
40°<br />
25°<br />
0°<br />
Humulinic<br />
acid<br />
100°<br />
40°<br />
25°<br />
0°<br />
Humulone<br />
Lupulone<br />
2 4 6 8 10<br />
pH<br />
Fig. 8.2 Solubilities of humulinic acid, humulone <strong>and</strong> lupulone (Spetsig, 1955).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The structure of the - <strong>and</strong> -acids have been verified by synthesis. Acylation of<br />
phloroglucinol (8.18) gives the parent phloracylphenone (8.22); phlorisovalerophenone<br />
(8.22a) for humulone <strong>and</strong> lupulone <strong>and</strong> phlorisobutyrophenone (8.22b) for cohumulone<br />
<strong>and</strong> colupulone, etc.<br />
This can be alkylated with 3-methyl-2-butenyl bromide (8.24) (isoprene hydrobromide,<br />
dimethylallyl bromide), which is prepared by the 1,4-addition of hydrogen bromide to<br />
isoprene (8.23).<br />
Alkylation of phloracylphenone (8.22) can give a mixture of one mono- (8.25), two di-<br />
(8.26 <strong>and</strong> 8.27), one tri- (8.28) <strong>and</strong> one tetra- (8.29) isoprenyl derivative.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The tri-substitued derivatives are the -acids (8.2), the di-substituted derivatives (8.26)<br />
the deoxy- -acids; these <strong>and</strong> the monosubstituted derivatives (8.25) have been found in<br />
hops but neither the di-substituted derivative (8.27) nor the tetra-substituted derivative<br />
(8.29, lupones) have been found to occur naturally. With one molecule of base <strong>and</strong> 3methyl-2-butenyl<br />
bromide the mono-substituted derivatives (8.25) can be obtained in<br />
good yield <strong>and</strong> in liquid ammonia the -acids (8.28) are obtained in up to 70% yield<br />
(Collins et al., 1971) but the synthesis of the the deoxy- -acids (8.26), <strong>and</strong> thus the -<br />
acidsismoredifficult.Intheoriginalsynthesisof( )-humulone(Riedl,1951)theoverall<br />
yield was only 5.7%. Here the deoxyhumulone was oxidized to the lead salt of ( )humulone<br />
with oxygen in the presence of lead (II) acetate.<br />
Slightly better yields of deoxy- -acids were obtained using the weakly basic ionexchange<br />
resin DeAcidite H-IP (OH form) (Collins <strong>and</strong> Laws, 1973) <strong>and</strong> by using 3methyl-buten-3-ol<br />
with boron trifluoride-etherate as the alkylating agent (Collins <strong>and</strong><br />
Shannon, 1973). Deoxy- -acids (8.26) have also been prepared by the irradiation of -<br />
acids with ultraviolet light (Fern<strong>and</strong>ez, 1967). Better yields of -acids from deoxy- -<br />
acids are found when the autoxidation is carried out in the presence of triethyl phosphite<br />
(Sigg-Grutter <strong>and</strong> Wild, 1974). Without this reducing agent the intermediate hydroperoxide<br />
was isolated. Even with the optimal yields reported (Pfenninger et al.,1975)<br />
synthetic racemic -acids are unlikely ever to be as cheap as the naturally produced<br />
enantiomers from the hop.<br />
8.2.2 Biosynthesis of the hop resins<br />
The biosynthesis of the hop resin (Fig. 8.3) within the plant is thought to follow asimilar<br />
route to the chemical synthesis. When CH3. 14 CO2Na was injected into aripening hop<br />
plant the labelling of the radioactive humulone suggested that the phloroglucinol nucleus<br />
was made up from three acetate units. The acyl side chains were derived from amino<br />
acids or intermediates in their biosynthesis. Thus humulone <strong>and</strong> lupulone come from a<br />
leucine metabolite, cohumulone <strong>and</strong> colupulone from avaline metabolite <strong>and</strong> the adanalogues<br />
from isoleucine (with aromatic amino acids the prenylflavanoids are formed,<br />
see Section 8.2.5). Transamination <strong>and</strong> decarboxylation of the amino acids leads to the<br />
Coenzyme Aesters of isovaleric, isobutyric <strong>and</strong> 2-methylbutyric acids (8.30). These are<br />
thought to react with three molecules of malonyl Coenzyme A (8.31) to give the<br />
polyketide (8.32) which with an enzyme similar to chalcone synthetase forms the<br />
phloracylphenone (8.22) (Fung et al., 1997). This with dimethylallyl pyrophosphate<br />
(8.34) <strong>and</strong> with the enzyme(s) prenyltransferase(s) forms the mono-prenyl derivative<br />
(8.25), the deoxy- -acids (8.26) <strong>and</strong>, probably, the -acids. Oxidation of the deoxy- -<br />
acids gives the required -acids (8.1).<br />
Sincethe1950sithasbeenthoughtthatdimethylallylpyrophosphate(8.34),theparent<br />
of the isoprenoids, terpenoids <strong>and</strong> steroids, was formed via mevalonic acid (8.35) (see<br />
Fig. 8.4 Pathway A) <strong>and</strong> that this was the only route to these compounds. Now, an<br />
alternative route, via 1-deoxyxyulose-5-phosphate (8.37), has been found (Eisenreich et<br />
al., 1998. Fig. 8.4 Pathway B). Goese et al., 1999) have shown that the isoprenyl- side<br />
chains of humulone are formed by this latter route <strong>and</strong> that the biosynthesis of humulone<br />
goes through asymmetrical intermediate. According to the new pathway, glyceraldehyde<br />
3-phosphate, produced from glucose by the normal Embden-Meyerhof-Parnas pathway<br />
(Fig. 12.7) condenses with `active acetaldehyde' (8.36), produced from pyruvate <strong>and</strong><br />
thiamine, to give D-1-deoxyxylulose 5-phosphate (8.37). This re-arranges, via 2-Cmethylerythrose<br />
(8.38) <strong>and</strong> 2-C-D-erythritol 4-phosphate (8.39) to isopentenyl pyropho-<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
R·CO·CO2H<br />
R·CO·S·CoA (8.30)<br />
+ 3HO2C·CH2·CO·S·CoA (8.31)<br />
R·CO·CH2·CO·CH2·CO·CH2·CO·S·CoA (8.32)<br />
OH<br />
CO·R<br />
HO OH<br />
(8.22)<br />
OH<br />
CO·R<br />
HO OH<br />
HO<br />
– _<br />
R·CH·CO2<br />
· +<br />
NH3<br />
(8.25)<br />
OH<br />
CO·R<br />
OH<br />
H2C<br />
H3C<br />
C·CH2·CH2·O<br />
O<br />
··<br />
P<br />
·<br />
OH<br />
O<br />
O<br />
··<br />
P<br />
·<br />
OH<br />
OH<br />
isopentenyl pyrophosphate (8.33)<br />
H2C<br />
H3C<br />
C CH·CH2·O<br />
O<br />
··<br />
P<br />
·<br />
OH<br />
O<br />
O<br />
··<br />
P<br />
·<br />
OH<br />
OH<br />
γ-Dimethylallyl pyrophosphate (8.34)<br />
sphate (8.33). This with an isomerase is converted into dimethylallyl pyrophosphate<br />
(8.34) which is thought to be the biological isoprenylating agent. Isopentenyl<br />
pyrophosphate (8.33) <strong>and</strong> dimethylallyl pyrophosphate (8.34) can condense together to<br />
form, first, geranyl pyrophosphate (8.88), the parent of the monoterpenes, <strong>and</strong> then<br />
farnesyl pyrophosphate (8.105), the parent of the sesquiterpenes, both of which are<br />
important constituents of the essential oil (see later). From the limited data available<br />
(Eisenreich et al., 1998) it appears that both the mevalonic acid (8.35) <strong>and</strong> the deoxyxyulose<br />
pathways (Fig. 8.4) are found in most higher plants. Steroids are mainly formed<br />
HO<br />
HO<br />
deoxy-α-acids (8.26) α-Acids (8.1)<br />
Fig. 8.3 Biosynthesis of the -acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
OH<br />
CO·R<br />
O
Pathway A<br />
CH3<br />
CH3<br />
O<br />
SCoA<br />
O<br />
SCoA<br />
O O<br />
acetyl CoA acetoacetyl CoA<br />
Pathway B<br />
CO2H<br />
CH3<br />
O<br />
–CO2<br />
HO<br />
by the mevalonate pathway but isoprene <strong>and</strong> the essential oil constituents arrive by the<br />
deoxy-xyulose pathway.<br />
8.2.3 Analysis of the hop resins<br />
Procedures for the estimation of the total resins, total soft resins, <strong>and</strong> hard resins, by<br />
difference, (see definitions above) in hops <strong>and</strong> hop products are given in Analytica-EBC<br />
but since it was found that the -acids are the most important brewing principles few<br />
brewers bother to measure the total <strong>and</strong> soft resin contents. The -acids were originally<br />
estimated gravimetrically as their lead salts but since the precipitate is soluble in excess<br />
of the lead acetate reagent, trials had to be made so that only the correct amount of<br />
CH3<br />
SCoA<br />
NADPH<br />
OH<br />
3 ATP<br />
–CO2<br />
O O<br />
– Pi<br />
HO<br />
H3C<br />
N<br />
S<br />
mevalonate<br />
(8.35)<br />
+<br />
CHO<br />
OH<br />
CH2OX<br />
pyruvate (8.36) GAP<br />
glyceraldehyde<br />
3-phosphate<br />
HO<br />
O OH<br />
OX<br />
2C-methylerythrose<br />
(8.38)<br />
HO<br />
NAD(P)H<br />
HO OH<br />
O<br />
HO<br />
SCoA<br />
O O<br />
4<br />
3<br />
5<br />
O<br />
HMG CoA<br />
2<br />
1<br />
IPP isopentenyl<br />
pyrophosphate (8.33)<br />
H<br />
X = H:2C-methyl-D-erythritol<br />
X = phosphate:<br />
2C-methyl-D-erythritol<br />
4-phosphate<br />
(8.39)<br />
OX<br />
O<br />
CH3<br />
O<br />
H<br />
OH<br />
CH2OX<br />
OPP<br />
SCoA<br />
X = H:D-1-deoxyxylulose<br />
X = phosphate:<br />
D-1-deoxyxylulose<br />
5-phosphate<br />
(8.37)<br />
IPP<br />
isopentenyl<br />
pyrophosphate<br />
(8.33)<br />
OPP<br />
Fig. 8.4 Isoprenoid biosynthesis: (A) via the mevalonate pathway, (B) via the glyceraldehyde<br />
3-phosphate/pyruvate pathway (Rohmer, 1998).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Conductance<br />
Lead acetate (ml)<br />
Fig. 8.5 Conductometric titration of -acids.<br />
reagent was used making it alengthy process. It was found that if the conductivity of a<br />
methanolic solution of the -acids was measured during titration with methanolic lead<br />
acetatesolutionitdidnotincreaseuntiltherewasanexcessofthereagent.Thusifregular<br />
aliquots (0.20ml) of the methanolic lead acetate reagent (2 or 4%) are added <strong>and</strong> the<br />
conductivitymeasuredaftereachaddition,agraphcanbedrawnwheretheintersectionof<br />
the two straight portions provides the endpoint (Fig. 8.5). The absolute value of the<br />
conductivity is not needed <strong>and</strong> the lead acetate reagent can be st<strong>and</strong>ardized by asimilar<br />
titration against 0.100 N sulphuric acid. The shape of the graph is different in the<br />
presence of other solvents, so, in the approved method, pyridine (1ml) is added to the<br />
titration. The conductivity may be plotted against the volume of the reagent on Cartesian<br />
coordinate paper (Fig. 8.5) or the resistance may be plotted directly on to reciprocal ruled<br />
paper. Since the reaction of lead acetate is not specific for -acids, the result is expressed<br />
as the Lead Conductance Value (LCV). However, with fresh hops the LCV is very<br />
similar to the -acid content but, as hops age on storage, oxidation products are formed<br />
which may react with lead acetate.<br />
The -acids are the only hop resins which show significant optical activity ([ ]D 20<br />
237ë in hexane) so they can be estimated in the soft resins by polarimetry but this<br />
method has not been officially adopted.<br />
Alderton et al. (1954) measured the light absorption of humulone <strong>and</strong> lupulone under<br />
both acid <strong>and</strong> alkaline conditions (Fig. 8.6). In particular, they measured in alkaline<br />
solution the Absorbance (A) at 275nm ( min for both humulone <strong>and</strong> lupulone), 325nm<br />
( maxforhumulone)<strong>and</strong>355nm( maxforlupulone)<strong>and</strong>producedregressionequationsto<br />
determine both humulone <strong>and</strong> lupulone. This method was adopted by the ABSC when:<br />
-acids, %ˆd … 51:56A355 ‡73:79A325 19:07A275†<br />
-acids, %ˆd …55:57A355 47:59A325 ‡5:10A275†<br />
where dis the dilution factor.<br />
These regression equations contain large multiplying factors so the procedure requires<br />
ahigh degree of precision in instrument calibration <strong>and</strong> the purity of the solvents. This<br />
was the first method to give values for the % -acids. The assumption that the solution of<br />
hop resins is abinary mixture of -<strong>and</strong> -acids with constant background absorption is<br />
probably true with fresh hops but not with deteriorated samples. Indeed, Likens et al.<br />
(1970) using this method proposed that the ratio<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Specific absorption coefficient (l/g cm)<br />
50<br />
25<br />
Humulone<br />
complex,<br />
acid<br />
Humulone<br />
complex,<br />
alkaline<br />
0<br />
200 250 300 350 400<br />
Wavelength (nm)<br />
A275=A325 ˆHop Storage Index (HSI)<br />
Lupulone,<br />
alkaline<br />
Lupulone,<br />
acid<br />
Fig. 8.6 Absorption spectra of lupulone <strong>and</strong> humulone complex in acidic (0.002N) <strong>and</strong> alkaline<br />
(0.002 N) methanol (Alderton et al., 1954). Copyright (1954) American Chemical Society.<br />
was found to increase from c. 0.24in fresh hops to c. 2.5in completely oxidized lupulin.<br />
The measurement has been adopted by the ASBC. For more precise analysis some<br />
method of separation is necessary before measurement. Methods based on Dowex <strong>and</strong><br />
Sephadexion-exchange resinshave nowbeenarchived<strong>and</strong>replacedbyaninternationally<br />
agreed HPLC method using a250 4mm, 5 mRP18 Nucleosil C19 column. The<br />
chromatogram usually shows four peaks: cohumulone, humulone + adhumulone,<br />
colupulone <strong>and</strong> lupulone + adlupulone. Other systems will resolve humulone <strong>and</strong><br />
adhumulonebutitremainsdifficult.Verzele <strong>and</strong>De Keukeleire(1991) describeHPLC of<br />
an ethanolic extract of hops using two coupled columns with diode array detection when<br />
sixty peaks could be recognized including 10 -acids, 10 iso- -acids, <strong>and</strong> 11 deoxy- -<br />
acids. Theoccurrence ofiso- -acids (2.7%)inanethanolicextract ofhopsisnoteworthy.<br />
8.2.4 Isomerization of the -acids<br />
As long ago as 1925 it was suggested that the hydrolysis of humulone (8.1) to humulinic<br />
acid (8.3) proceeded via an intermediate (8.40, Fig. 8.7). Later this structure (8.40) was<br />
given to the bitter-tasting oil obtained by boiling humulone with N/15 (0.067M) sodium<br />
hydroxide solution for three minutes (Windisch et al., 1927). It was originally called<br />
`Resin A' but later the name `isohumulone' was adopted for the parent ofthe iso- -acids.<br />
The iso- -acids are much more soluble in water (c. 120mg/l) than the -acids (3mg/l)<br />
<strong>and</strong> according to Peacock (1998) nine times more bitter. To account for the formation of<br />
isobutyraldehyde (8.14), asecond pathway leading to 8.14 <strong>and</strong> `Resin B' (4-acetylhumulinic<br />
acid, 8.42) was proposed. Later the products of this second pathway were<br />
thought to be formed via allo-iso- -acids (8.41). However, only 4.5% of allo-iso- -acids<br />
were formed inboiling wort but better yields were found at pH9.0. Less than 1mg/l were<br />
found in beer. So the iso- -acids, isohumulone, isocohumulone <strong>and</strong> isoadhumulone<br />
(8.40) are the major bittering principles in beer.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
HO<br />
OH<br />
CO·R<br />
O<br />
HO<br />
They are usually estimated by the light absorption of a solvent extract. In the<br />
internationally agreed method degassed, acidified beer (10 ml) is extracted with isooctane<br />
(2, 2, 4-trimethylpentane) (20 ml) <strong>and</strong>, after centrifugation, the absorbance of the<br />
isooctane layer is read at 275 nm in a 1 cm cell against an isooctane blank. This is a<br />
modification of the method of Moltke <strong>and</strong> Meilgaard (1955) who gave a regression<br />
equation to present results as `isohumulones' mg/l. In order to avoid assumptions about<br />
the chemical nature of the bittering principles in beer, the EBC Analysis Committee<br />
simplified the regression equation <strong>and</strong> gave the results in Bitterness Units (BU):<br />
Bitterness Units (BU) = 50 absorbance<br />
CO·R CO·R<br />
CO CO<br />
HO OH<br />
(8.1) humulone (8.40) isohumulone (8.41) allo -isohumulone<br />
H3C<br />
H3C<br />
HO<br />
CO·R CO·R<br />
With worldwide adoption they are sometimes called International Bitterness Units (IBU).<br />
Most commercial beers contain 10±50 BU but exceptional beers with up to 100 BU have<br />
been reported (Glaser, 2002). Rigby <strong>and</strong> Bethune (1955) estimated the iso- -acids in an<br />
isooctane extract of beer by measuring the light absorption at 255 nm after dilution with<br />
alkaline methanol. Beers produced by conventional wort boiling contain only traces of -<br />
acids or humulinic acids which would interfere with the above two methods. However,<br />
worts, isomerized hop extracts <strong>and</strong> beers brewed therefrom may contain such impurities<br />
so accurate values for the iso- -acids can be obtained only after a suitable separation<br />
technique. Originally countercurrent distribution was the method of choice but HPLC is<br />
more likely to be used today.<br />
O<br />
O<br />
(8.3) humulinic acid<br />
OH<br />
OH HO<br />
OH<br />
H CO·CH3<br />
O<br />
O<br />
(8.42) 4-acetylhumulinic acid<br />
+ +<br />
H3C<br />
C CH·CH2·CO2H CH·CHO<br />
H3C<br />
(8.13) isohexenoic acid (8.14) isobutyraldehyde<br />
Fig. 8.7 Isomerization <strong>and</strong> hydrolysis of humulone (R ˆ Bu i without stereochemistry).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Having established that the -acids are isomerized into iso- -acids during wort<br />
boiling <strong>and</strong> that the iso- -acids are the main bittering principles in beer, hop utilization<br />
can be defined as:<br />
%Hop utilization ˆ<br />
Amount of iso- -acids in beer<br />
Amount of -acids in hops used<br />
In conventional wort boiling only about 50% of the -acids available in the hops go into<br />
solution <strong>and</strong> further losses on to the break <strong>and</strong> on to the yeast during fermentation occur<br />
so that the overall utilization into beer will seldom exceed 40% <strong>and</strong> may be as low as<br />
10%.Inwortboilinghigherutilizationoccurswithweakworts<strong>and</strong>lowlevelsofhopping.<br />
Indeed, the solubility of humulone (<strong>and</strong> other -acids) is alimiting factor in utilization<br />
(Fig. 8.2). Only 50±60% of pure humulone was isomerized during a1.5hboil while with<br />
the same amount of -acids in hops 65±75% were utilized in the same period. Pure<br />
humulone in boiling wort forms an oily layer or droplets of minimal surface area <strong>and</strong> the<br />
utilization is improved when the resin is spread over an inert surface such as hops, break,<br />
or even Celite. Utilization is also improved at higher pH values but only small variations<br />
are possible in wort. However, if the -acids are isolated from the hops, isomerized at<br />
higher pH values <strong>and</strong> then added to the beer after fermentation much better utilization is<br />
obtained.<br />
Many brewers now use such isomerized extracts. Some brewers will use low levels of<br />
hops in the copper, with improved utilization, <strong>and</strong> then achieve the desired bitterness by<br />
addition of isomerized extracts after fermentation, others will depend entirely on postfermentation<br />
additions. The isomerized extract must not contain any -acids as they<br />
would be precipitated in the beer necessitating a further filtration with loss of iso- -<br />
acids. Many patents exist for the preparation of isomerized extracts. For example, the -<br />
acids may be extracted from a solvent (ethanol or CO2) extract of hops using either<br />
disodium or dipotassium carbonate <strong>and</strong> the carbonate solution boiled to effect<br />
isomerization. The weaker -acids will not dissolve in the carbonate solutions.<br />
Alternatively, both - <strong>and</strong> -acids may be extracted with alkali hydroxides <strong>and</strong> the<br />
solution then saturated with CO 2 gas to lower the pH <strong>and</strong> precipitate the -acids. Care<br />
must be taken that the -acids are not boiled with alkali hydroxides as hydrolysis to<br />
humulinic acids will occur. The iso- -acid solution is best injected into a beer main<br />
because addition to conditioning tanks may cause local supersaturation <strong>and</strong> precipitation<br />
of the iso- -acids which will only slowly redissolve. The calcium <strong>and</strong> magnesium salts<br />
of the -acids can be isomerized by heating at 70 ëC for two hours. The iso- -acid salts<br />
formed are finely ground (particles < 10 m) <strong>and</strong> added to conditioning tanks when at<br />
least 24 hours are necessary to achieve solution <strong>and</strong> 85% utilization. The -acids<br />
recovered from the above processes may be added to the copper with or without<br />
deliberate oxidation. The official methods of analysis give two HPLC methods to<br />
determine the iso- -acids in isomerized hop extracts. Both allow for the separation of<br />
isocohumulone, isohumulone <strong>and</strong> isoadhumulone; one uses 4-methylbenzophenone as<br />
internal st<strong>and</strong>ard, the other -phenylchalcone. The gross structure of the iso- -acids has<br />
been confirmed by synthesis (Ashurst <strong>and</strong> Laws, 1966, 1967) but the low yield means<br />
that this is not a practical route to synthetic iso- -acids.<br />
The structures assigned to isohumulone (8.40) <strong>and</strong> humulinic acid (8.3) both contain<br />
two chiral centres so each should exist as two pairs of enantiomers. However, since<br />
natural (R)( )-humulone is a single enantiomer only two diasteroisomeric forms are<br />
found, the cis- <strong>and</strong> trans-isomers (see definitions above). The enantiomers of these<br />
compounds would be obtained from unnatural (S)(+)-humulone. Countercurrent<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
100
distribution of the bittering substances in beer showed three peaks corresponding to<br />
isocohumulone, isohumulone <strong>and</strong> isoadhumulone but each peak was broader than that<br />
calculated for apure compound <strong>and</strong> two theoretical curves could be fitted under each<br />
observed peak. The same broadened pattern was exhibited by isohumulone obtained by<br />
boiling humulone in N/15 sodium hydroxide for three minutes, 0.1N-disodium carbonate<br />
for 30min. or in apH5.0 buffer solution. The proton magnetic resonance spectra of these<br />
isohumulone preparations were in agreement with them being a mixture of two<br />
stereoisomers (Burton et al., 1964). Eventually the two stereoisomers of isohumulone<br />
were separated by reversed-phase partition chromatography (Spetsig, 1964), countercurrent<br />
distribution after 2000 transfers (Alderweireldt et al., 1965), partition<br />
chromatography on silica gel (Clarke <strong>and</strong> Hildebr<strong>and</strong>, 1965) <strong>and</strong>, later, thin layer<br />
chromatography (Aitken et al., 1970) <strong>and</strong> shown to be the cis- (8.43) <strong>and</strong> trans- (8.44)<br />
-isomers as expected. At the same time it was found that irradiation ofhumulone at either<br />
365 or 254nm gave pure crystalline trans-isohumulone (8.44 photoisohumulone) (Clarke<br />
<strong>and</strong>Hildebr<strong>and</strong>,1965(seealsoSharpe<strong>and</strong>Ormrod,1991)),sothisisomerismorereadily<br />
available pure than the oily cis-isomer.<br />
Chemically, the isomerization of humulone is atype of benzilic acid or acyloin<br />
rearrangement <strong>and</strong> the mechanism is given in Fig. 8.8. The isomerization follows firstorder<br />
kinetics in buffer solutions of constant pH but falls off in wort probably due to the<br />
lowering of the pH. The two isomers of isohumulone are not readily converted into each<br />
other but, since the isomerization of humulone is reversible, they can be interconverted<br />
via humulone. When isohumulone is heated alone, in wort, in abuffer solution pH4.5, or<br />
in 0.1N-disodium carbonate, 10±15% of humulone is formed. The same percentage of<br />
humulone is found when isohumulone is shaken in atwo phase system of isooctane <strong>and</strong> a<br />
pH5.0 buffer solution so it follows that pure isohumulone cannot be isolated by<br />
countercurrent distribution. Koller (1969) found that the isomerization of the -acids was<br />
catalysed by divalent ions, especially calcium <strong>and</strong> magnesium, <strong>and</strong> that the iso- -acids<br />
prepared in this way were free of humulinic acids. As mentioned above, the calcium <strong>and</strong><br />
magnesium salts of the iso- -acids can be used as bittering agents <strong>and</strong> they are formed in<br />
the preparation of isomerized pellets (Chapter 7).<br />
The separation <strong>and</strong> analysis of the six iso- -acids, as in isomerized extracts <strong>and</strong> beer,<br />
remains difficult; HPLC is the method of choice. Thornton et al. (1993) found that the<br />
trans-iso- -acids in ethyl acetate formed insoluble salts with dicyclohexylamine leaving<br />
the cis-isomers in solution. Thus amixture of six iso- -acids is converted into two<br />
mixtures of three which are easier to resolve by HPLC (Hughes, 1996). Using these<br />
techniques Hughes et al. (1997) were able to show that the amount of cis- <strong>and</strong> trans-iso-<br />
-acids formed depends on the reaction conditions employed (Table 8.2) <strong>and</strong> that the cis<strong>and</strong><br />
trans-isomers behave differently. During wort boiling with different hop products<br />
(Type 90 pellets from Wye Target <strong>and</strong> a liquid CO2 extract from Galena) the proportion<br />
of isocohumulone present is at a maximum after 30 min.; the utilization of humulone <strong>and</strong><br />
adhumulone is slower. This may be due to the fact that in wort cohumulone (pKa 4.7) is<br />
more ionized than humulone (pKa 5.5).<br />
The cis/trans-ratio of the iso- -acids (68:32) appears to be constant during wort boiling<br />
<strong>and</strong> it is generally thought that this ratio represents the thermodynamic equilibrium, i.e., the<br />
cis-isomers are energetically more likely to be formed during wort boiling. During<br />
fermentation the iso- -acids are lost by adsorption on to the yeast head. The less polar iso-<br />
-acids, isohumulone <strong>and</strong> isoadhumulone, interact more strongly with the yeast cells<br />
causing an enrichment of isocohumulone in the beer. Earlier workers had found that<br />
cohumulone was utilized into beer better than the other -acids but the above work<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
HO<br />
H<br />
HO<br />
O<br />
(8.1a)<br />
H<br />
O O<br />
O<br />
O O<br />
H<br />
O<br />
H O<br />
B –<br />
O – O<br />
H O<br />
HO<br />
O –<br />
suggests that this is due to the preferential removal of the other iso- -acids. The importance<br />
of the proportion of isocohumulone in the beer iso- -acids on the flavour of beer is still a<br />
subject for debate. Rigby (1972) said that isocohumulone had a harsh bitter flavour <strong>and</strong>, in<br />
the days of boiling hops in wort, brewers preferred hops with a low proportion of<br />
cohumulone in their -acids. This harsh flavour has not been commented upon by later<br />
O<br />
H +<br />
O<br />
O<br />
O O<br />
OH<br />
H<br />
HO<br />
O<br />
OH<br />
(8.43a) cis (8.44a) trans<br />
Fig. 8.8 Mechanism for the isomerization of humulone (De Keukeleire <strong>and</strong> Verzele, 1971).<br />
Copyright (1971) with permission from Elsevier.<br />
Table 8.2 Typical ratios of cis/trans-iso- -acids under a range of isomerization conditions<br />
(Hughes et al., 1997)<br />
Means of isomerization cis-isomers trans-isomers Reference<br />
(%) (%)<br />
Magnesium oxide 80 20 Hughes (unpublished)<br />
Wort boiling 68 32 Verzele <strong>and</strong> De Keukeleire (1991)<br />
Aqueous alkali 55 45 Koller (1969)<br />
Light 0 100 Clarke <strong>and</strong> Hildebr<strong>and</strong> (1965)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H +
workers. Of the four major iso- -acids found in beer cis-isohumulone was found to be the<br />
most bitter. (c. 1.82 times more bitter than trans-isohumulone), trans-isocohumulone was<br />
the least bitter (c. 0.74 that of trans-isohumulone) <strong>and</strong> there was little difference in<br />
bitterness between cis-isocohumulone <strong>and</strong> trans-isohumulone (Hughes <strong>and</strong> Simpson,<br />
1996). Hughes (2000) has reviewed the significance of the iso- -acids for beer quality.<br />
It is well known that beer deteriorates on storage, slowly at 0 ëC <strong>and</strong> in cool cellars,<br />
but more rapidly in centrally heated houses <strong>and</strong> supermarkets. Forced ageing at 37 ë or<br />
40 ëC is used to forecast beer stability. During the storage of beer there is a loss of iso- -<br />
acids probably due to autoxidation <strong>and</strong> free radical reactions accelerated by iron ions <strong>and</strong><br />
hydrogen peroxide (Kaneda et al., 1992). It follows that the level of antioxidants in the<br />
beer will moderate the rate of deterioration. However, De Cooman et al. (2000) found<br />
that both in lagers <strong>and</strong> top-fermented beers the trans-iso- -acids deteriorated at a much<br />
faster rate than the cis-isomers. During the first 15 months of storage the loss of iso- -<br />
acids was mainly due to decomposition of the trans-isomers. In the trans-iso- -acids the<br />
double bonds in the 3-methyl-2-butenyl- <strong>and</strong> 4-methyl-3-pentenoyl-side chains lie close<br />
together <strong>and</strong> it is suggested that this increases their susceptibility to autoxidation. As<br />
mentioned earlier, the - <strong>and</strong> -acids also undergo autoxidation. The -acids, with<br />
geminal-isoprenyl groups, are too sensitive to autoxidation to be separated by<br />
countercurrent distribution. The unsaturated side-chains of the hop resins are<br />
undoubtedly the site of attack at the carbon next to the double bond <strong>and</strong> the<br />
hydroperoxides formed could break down to isobutyraldehyde (8.14) <strong>and</strong> acetone (Fig.<br />
H3C<br />
H3C<br />
H3C<br />
H3C<br />
H3C<br />
H3C<br />
C<br />
C C<br />
CH·C<br />
C<br />
H<br />
O<br />
H<br />
C<br />
H<br />
•<br />
CH·R<br />
H R<br />
H3C<br />
H3C<br />
O·O·H<br />
C C<br />
–H•<br />
O2<br />
H3C<br />
H3C<br />
H<br />
CH2·R<br />
H3C<br />
H3C<br />
H·O·O<br />
H3C<br />
H3C<br />
•C C<br />
C C<br />
H<br />
C O + O C<br />
(8.14) isobutyraldehyde (8.15) acetone<br />
CH·R<br />
H<br />
CH·R<br />
H<br />
CH2·R<br />
Fig. 8.9 Proposed autooxidation of 3-Methylbut-2-enyl side chains.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
HO<br />
OH<br />
CO·R<br />
O<br />
HO<br />
HO<br />
(8.1) α-Acids (8.45) dihydro-α-acids (8.51) dehydro-α-acids<br />
HO<br />
CO<br />
O<br />
HO<br />
HO<br />
OH<br />
CO·R<br />
(8.9) tetrahydro-α-acids (8.52)<br />
(8.40) iso-α-acids (8.47) tetrahydoiso-α-acids<br />
O<br />
HO<br />
H·C·OH<br />
(8.49) ρ-Iso-α-acids<br />
CO·R<br />
OH<br />
CO·R<br />
O<br />
HO<br />
CO<br />
8.9).Thisroutetoisobutyraldehyde isprobablymoreimportantthanthesecondpathway<br />
shown in Fig. 8.7.<br />
Obviously if these unsaturated side-chains are reduced (Fig. 8.10), the saturated<br />
products will be much less sensitive to autoxidation. Hydrogenation of humulone (8.1a)<br />
in the presence of platinum (IV) oxide give first the dihydrohumulone (8.45a) <strong>and</strong> then<br />
tetrahydrohumulone (8.9a). The alternative dihydrohumulone (8.46a) does not appear to<br />
O<br />
O<br />
OH<br />
HO<br />
H·C·OH<br />
CO·R CO·R<br />
O<br />
O<br />
HO<br />
OH<br />
O<br />
HO<br />
CO·R<br />
OH<br />
CO·R<br />
H3C<br />
H3C<br />
CO·R<br />
C CH·CH2·SH<br />
(8.48) 3-Methyl-2-butenyl<br />
mercaptan (thiol)<br />
(8.50) hexahydroiso-α-acids (8.46)<br />
Fig. 8.10 Reduced iso- -acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
OH<br />
HO<br />
HO<br />
O<br />
OH<br />
CO·R<br />
O
e formed. As mentioned, hydrogenation of humulone in the presence of palladium<br />
chloride leads to hydrogenolysis loss of an isoprenyl side-chain (Fig. 8.1). Similarly<br />
hydrogenolysis of the -acids (8.2) gives tetrahydrodeoxy- -acids (8.8) which can be<br />
oxidized to tetrahydro- -acids (8.9). The tetrahydro- -acids were resolved by countercurrent<br />
distribution <strong>and</strong> these reactions were used to determine the analogues present in<br />
the -acidswhenitwasfoundthatthe -acidsarealwaysricherintheco-componentthan<br />
the -acids. So tetrahydro- -acids, obtained from the -acids will always be richer in<br />
cohumulone thanthoseobtainedbyhydrogenationofthe -acids. Inaddition,tetrahydro-<br />
-acids obtained by hydrogenation of -acids will be optically active, while those from<br />
-acids will be racemic.<br />
Both the dihydro- -acids <strong>and</strong> the tetrahydro- -acids can be isomerized to the<br />
corresponding iso- -acids. In particular, tetrahydroiso- -acids (8.47) are light stable,<br />
more bitter than the unsaturated iso- -acids <strong>and</strong> more potent foam stabilizers (Baker,<br />
1990). Thetetrahydroiso- -acids areapprovedby the FDAfor useasbitteringagents<strong>and</strong><br />
many brewers use them, with or without iso- -acids, for post-fermentation bittering. The<br />
tetrahydroiso- -acids can be prepared from -acids either by hydrogenation followed by<br />
isomerization or by isomerization followed by reduction. The latter route is preferred <strong>and</strong><br />
the two stages can be combined to give an efficient one-step preparation (Hay <strong>and</strong><br />
Homiski,1991).Numerouspatentsalsodescribethepreparationoftetrahydroiso- -acids.<br />
Analysis of beer bittered with amixture of iso- -acids <strong>and</strong> tetrahydroiso- -acids showed<br />
the expected 12 peaks on the HPLC chromatogram. After 12 months storage at 25ëC,<br />
only the trans-iso- -acids had deteriorated significantly; the tetrahydroiso- -acids were<br />
stable (De Cooman et al., 2000). It is noteworthy that these workers developed aHPLC<br />
system to resolve the 12 peaks.<br />
Beforetheuseoftetrahydroiso- -acids(8.47)asbitteringagentsitwasfoundthatbeer<br />
stored in clear glass bottles <strong>and</strong> exposed to sunlight developed an unpleasant skunky,<br />
sunstruck flavour found to be due to 3-methyl-2-butenyl mercaptan (thiol) (8.48). It was<br />
envisagedthatphotolysisoftheiso- -acidseitherproduceda3-methyl-2-butenyl-radical<br />
directly or produced a4-methyl-3-pentenoyl- radical which decarbonylated to the 3methyl-2-butenyl-<br />
radical. This radical then scavenges athiol group from any available<br />
sulphur amino acid orprotein (see also Heyerick et al., 2003). It was found that reduction<br />
of the iso- -acids with sodium borohydride produced abittering agent insensitive to<br />
light, theso-called rho-( )-iso- -acids (8.49).Borohydride reduction attacks the carbonyl<br />
group of the 4-methyl-3-pentenoyl side chain of the iso- -acids forming asecondary<br />
alcohol <strong>and</strong> anew chiral centre. Each iso- -acid will form two -iso- -acids <strong>and</strong> all four<br />
isomers of -isohumulone have been separated <strong>and</strong> characterized. They are less bitter<br />
than normalisohumulonebutnosignificantdifference wasnoticed between thebitterness<br />
of the individual isomers. Sodium borohydride reduction of tetrahydroiso- -acids (8.47)<br />
produces hexahydroiso- -acids (8.50). Again each steriosomer of the tetrahydroisohumulone<br />
will produce two isomeric hexahydrohumulones. The hexahydroiso- -acids are<br />
light stable, more bitter than conventional iso- -acids but less bitter than the<br />
tetrahydroiso- -acids <strong>and</strong> lead to an unnaturally dense foam. The properties of these<br />
semi-synthetic bittering agents are compared in Table 8.3. The dicyclohexylamine salts<br />
of the iso- -acids , the -iso- -acids, <strong>and</strong> the hexahydroiso- -acids have been prepared<br />
as st<strong>and</strong>ards for HPLC analysis (Maye et al., 1999); the tetrahydroiso- -acids may be<br />
used directly after recrystallization.<br />
The dihydrohumulone (8.45a) has been found in hop extracts, especially those<br />
containing the water-soluble fraction (Moir <strong>and</strong> Smith, 1995). The level decreases during<br />
storage <strong>and</strong> the formation is catalysed by the monovalent sodium <strong>and</strong> potassium ions. It<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 8.3 Semi-synthetic bittering agents (Marriott, 1999)<br />
Product Relative bitterness Relative foam enhancement Light stable<br />
at equivalent bitterness<br />
Iso- -acids 1.0 x No<br />
Rho ( )-iso- -acids 0.65 x Yes<br />
Tetrahydroiso- -acids 1.7 xxx Yes<br />
Hexahydroiso- -acids 1.1 xxxx* Yes<br />
* Foam unnaturally dense<br />
was found that at low temperatures, e.g., 3ëC, adisproportionation reaction occurs: two<br />
moleculesofhumulonegive onemoleculeofdihydrohumulone(8.45a)<strong>and</strong>onemolecule<br />
of dehydrohumulone (8.51a). At higher temperatures, e.g., 35ëC, the isomerization of the<br />
-acids predominates. Dehydrohumulone (8.48) is avery reactive compound which may<br />
either cyclize to apyran (8.52) or polymerize.<br />
8.2.5 Hard resins <strong>and</strong> prenylflavonoids<br />
By definition the hard resin is soluble in methanol <strong>and</strong> diethyl ether but insoluble in<br />
hexane. As hops age during storage the level of soft resins falls <strong>and</strong> that of the hard<br />
resins increases. Thus we must distinguish between `native' hard resins, present in<br />
fresh hops, <strong>and</strong> those formed by oxidation during storage. Many of the native hard<br />
resins are made up of prenylflavonoids which are deposited with the soft resins <strong>and</strong><br />
essential oils in the lupulin gl<strong>and</strong>s. Hop flavonoids have been reviewed by Stevens et<br />
al. (1998) including, as well as the prenylflavonoids, flavonoid glycosides, condensed<br />
tannins <strong>and</strong> other polyphenols which are largely soluble in water. The biosynthesis of<br />
the flavonoids (Fig. 8.11) is similar to that of the hop resins. The aromatic amino acids<br />
phenylalanine <strong>and</strong> tyrosine (8.53) lose ammonia to give, for example, p-coumaric acid<br />
(8.54), the Coenzyme Aester of which condenses with three molecules of malonyl<br />
Coenzyme Ato form achalcone (8.55, chalconaringenin) which can cyclize to a<br />
flavanone (8.56, naringenin). Dehydrogenation of flavanones can give flavones while<br />
dehydrogenation <strong>and</strong> hydroxylation leads to flavanols, e.g., kaempferol (8.57) <strong>and</strong><br />
quercetin (8.58).<br />
The major component of the prenylflavonoids in hops (Fig. 8.12) is the chalcone<br />
xanthohumol, first isolated in 1913 although the structure, 6'-O-methyl-3'-prenylchalconaringenin<br />
(8.59), was not worked out until the 1960s. It is accompanied by smaller<br />
amounts of the related flavanone isoxanthohumol (8.65, humulol, 5-O-methyl-8prenylnaringenin).<br />
This is racemic <strong>and</strong> it is thought that the lupulin gl<strong>and</strong>s lack the<br />
enzyme chalcone isomerase which converts chalcones to flavanones. Accordingly, all the<br />
racemic naringenin derivatives isolated from hops <strong>and</strong> beer are thought to be artefacts.<br />
Optically active isoxanthohumol has been isolated from Sophora angustifolia. Hops<br />
contain 0.1±0.8% of xanthohumol, the level falling on storage. Part of a growth of<br />
Eastwell Golding hops was freeze dried when the level of xanthohumol was 0.86%.<br />
When the same hops were kilned, with or without sulphur, the level fell to 0.31%. The<br />
level of isoxanthohumol appeared to be the same in all three samples. Stevens et al.<br />
(1997) found 8 prenylflavonoids <strong>and</strong> 3'-geranylnaringenin (8.61) in hops. Xanthohumol<br />
accounted for 82±89% of this fraction, desmethylxanthohumol (8.60) 2±3%, dehydrocycloxanthohumol<br />
(8.63) 2±4%, <strong>and</strong> dehydrocycloxanthohumol hydrate (8.64) 3±5%.<br />
The level of isoxanthohumol (8.65) was only 1±2% but during the brewing process nearly<br />
all of the chalcones are cyclized to flavanones. Xanthohumol can only cyclize to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
OH<br />
–<br />
HO CH2·CH·CO2<br />
O<br />
+<br />
NH3<br />
HO OH<br />
OH<br />
–NH3<br />
HO CH CH·CO2H<br />
(8.53) tyrosine (8.54) p-Coumaric acid<br />
OR<br />
O<br />
(8.57) kaempferol (R = H)<br />
HO<br />
isoxanthohumol but desmethylxanthohumol (8.60) gives a mixture of 6- (8.66) <strong>and</strong> 8prenylnaringenin<br />
(8.67). Similarly, 3'-geranylchalconaringenin (8.61) can give two<br />
products. 6-Geranylnaringenin (8.68) has been found in beer but not in hops so it is<br />
presumably formed from (8.61) during the brewing process.<br />
During the brewing process 69% of the available xanthohumol was in solution after<br />
the whirlpool <strong>and</strong> 13% was recovered from the spent hops. However, further losses on to<br />
the trub <strong>and</strong> the yeast mean that only about 30% of the available xanthohumol ends up in<br />
the beer (as isoxanthohumol). Thus the level of xanthohumol in 11 beers was 0.002±<br />
0.69 mg/l while that of isoxanthohumol was 0.04±3.44 mg/l. Similarly, only 11% of the<br />
available desmethylxanthohumol was found in beer as a mixture of 6- <strong>and</strong> 8prenylnaringenin<br />
(Stevens et al., 1999). The prenylflavonoids show interesting biological<br />
activities in vitro including antiproliferative effects on cancer cells, anti-carcinogenic<br />
C<br />
CH<br />
O<br />
(8.55) chalconaringenin<br />
OH<br />
OH<br />
O<br />
O<br />
(8.56) naringenin<br />
CH OH<br />
HO<br />
OH<br />
+<br />
3 HO2CH2·CO·S·CoA (8.31)<br />
OH<br />
Fig. 8.11 Biosynthesis of flavonoids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
O<br />
OR<br />
(8.58) quercetin (R = H)<br />
OH<br />
OH
R2<br />
HO 4'<br />
5'<br />
O·R3<br />
6'<br />
1'<br />
R1 3'<br />
2'<br />
OH O<br />
R1 R2 R3<br />
Prenyl H Me xanthohumol (8.59)<br />
Prenyl H H desmethylxanthohumol (8.60)<br />
Geranyl H H 3'-Geranylchalconaringenin (8.61)<br />
Prenyl Prenyl Me 5'-Prenylxanthohumol (8.62)<br />
HO<br />
O<br />
O·Me<br />
(8.63) dehydrocycloxanthohumol<br />
OH<br />
4<br />
(8.64) dehydrocycloxanthohumol hydrate<br />
HO<br />
R2<br />
O<br />
6<br />
7<br />
5<br />
OH<br />
R3<br />
8<br />
OH<br />
O<br />
O<br />
A C<br />
OR1<br />
O<br />
O·Me<br />
R1 R2 R3<br />
Me H Prenyl isoxanthohumol (8.65)<br />
H Prenyl H 6-Prenylnaringenin (8.66)<br />
H H Prenyl 8-Prenylnaringenin (8.67)<br />
H Geranyl H 6-Geranylnaringenin (8.68)<br />
effects, effects on lipid metabolism, oestrogenic <strong>and</strong> antimicrobial activity (Stevens, et<br />
al., 1998). Whether the concentrations in beer are sufficient to influence the consumer is<br />
yet to be determined. For example, the highest concentration of 8-prenylnaringenin,<br />
reported to be a phytoestrogen, found in beer was 19.8 M/litre.<br />
O<br />
2<br />
3<br />
OH<br />
Fig. 8.12 Prenylflavonoids in hops.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
B<br />
OH<br />
OH
8.2.6 Oxidation of hop resins<br />
As mentioned above hops deteriorate on storage, largely by oxidation; the soft resins<br />
decrease, the hard resins increase. The rate of deterioration can be followed by the Hop<br />
Storage Index (HSI) (p. 269). Nikerson <strong>and</strong> Likens (1979) found the regression equation<br />
%… ‡ † lost = 110 log (HSI/0.25)<br />
<strong>and</strong> that there was a linear relationship between the %( + ) lost, determined from the<br />
HSI, <strong>and</strong> the formation of hard resin. The rate of deterioration depends very much on the<br />
variety which appears to determine the length of the lag phase before deterioration starts.<br />
Thereafter the loss of both - <strong>and</strong> -acids can be fitted to either zero or first order kinetic<br />
equations (Green, 1978). To minimize such deterioration brewers used to keep their hops<br />
in cold store (0.20 ëC) which was expensive for such a bulky crop. Whitear (1965, 1966)<br />
questioned the need for this with copper hops since he found the bittering value of aged<br />
hops did not decline as rapidly as the analyses indicated (Fig. 8.13). However, on a<br />
commercial scale hops lost as much as 15±20% of the brewing value of the original -<br />
acids over two years at ambient temperature. From these <strong>and</strong> other trials, it was suggested<br />
that the level of -acids at harvest was the best guide to the amount of hops to be used in<br />
the copper. Accordingly, the bulk of the English hop crop was analysed within a month of<br />
harvest (see J. Inst. <strong>Brewing</strong>, 1969±1988). The production of hop pellets greatly reduced<br />
the volume of the crop <strong>and</strong> the saving in cold storage space helped to offset the cost of<br />
pelletization. Further, the pellets could be stored in an inert atmosphere more easily than<br />
bales or pockets.<br />
Old hops develop a cheesy aroma due to isovaleric, isobutyric <strong>and</strong> 2-methylbutyric<br />
acids produced by oxidative cleavage of the acyl side chains of the resins. Fresh hops<br />
contain 1±3% of volatile acids; after three years storage this increased to 20%. As<br />
mentioned earlier photolysis of colupulone (8.2b) causes loss of a 3-methyl-2-butenyl<br />
(isoprenyl) side chain to give deoxycohumulone (8.26b). Similarly, mild acid hydrolysis,<br />
wort boiling or atmospheric degradation of the -acids causes loss of isoprenyl side<br />
chains giving eventually the phloracylphenone (8.22). The isoprenyl side chains so<br />
displaced form mainly 2-methyl-3-buten-2-ol (8.69) which is absent from green hops but<br />
found after kilning <strong>and</strong> increases in stored hops. It is thought to be responsible for the<br />
soporific effects of hops <strong>and</strong> their use in hop pillows. Other volatile compounds which<br />
develop during hop storage include 2-methyl-2-butene, isoprene (8.23), 3-methyl-2-<br />
α-Acid content<br />
A<br />
Time<br />
B Bittering value<br />
α-Acids<br />
Lead acetate<br />
values<br />
Fig. 8.13 Schematic diagram of the changes in resin content <strong>and</strong> bittering value of hops during<br />
storage (Whitear, 1965, 1966).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
uten-1-ol, 5,5-dimethyl-(5H)-2-furanone (8.70), acetone, methyl isopropyl ketone, <strong>and</strong><br />
methyl isobutyl ketone. The last two compounds are thought to be relics of the acyl side<br />
chains.<br />
As mentioned above, isoprenyl side chains are very sensitive to autoxidation but the<br />
ring system can also be attacked <strong>and</strong> such reactions can be studied in saturated<br />
compounds produced by hydrogenation. Many compounds have been characterized by<br />
deliberate oxidation of individual hop resins (Fig. 8.14) but it is often difficult to find<br />
these oxidation products in stored hops or beers brewed therefrom. For example,<br />
oxidation of humulone (8.1a) with organic peroxides in the presence of base gives transhumulinone<br />
(8.71), pKa 2.8, which is intensely bitter, but, over 50 years, its presence in<br />
stored hops has not been established unequivocally. In contrast, oxidation of humulone<br />
with lead tetra-acetate gives tricyclodehydroisohumulone (8.72, TCD), which has been<br />
found in stored hops, in amounts up to 0.3%, <strong>and</strong> in beer (4 ppm). The bitterness of TCD<br />
is reported to be 70% that of trans-isohumulone (8.44) <strong>and</strong> it may contribute as much as<br />
5% of the total bitterness of beer. Oxidation of humulone with monoperphthalic acid<br />
gives (8.73) which may account for 1±5% of the hard resin. This compound (8.73) may<br />
bean intermediate inthe formationof(8.74)obtainedby the autoxidationofhumulone in<br />
hexane. (8.74) has abitter taste <strong>and</strong> is more water soluble than most hop resins but the<br />
bitterness is lost on boiling. Oxidation of humulone with m-chloroperbenzoic acid gives<br />
(8.75) but the bitterness <strong>and</strong> importance of this compound is not reported. Boiling<br />
humulone in aqueous buffer solutions gives, in addition to isohumulone, acomplex<br />
mixture of products from which (8.76) has been isolated. Later, (8.76a) <strong>and</strong> (8.76b) were<br />
detected in beer but only in trace amounts. When oxygen is bubbled through aboiling<br />
solution of humulone containing Celite aseries of abeo-isohumulones (8.77±8.83 Fig.<br />
8.15) are formed. They are reported to be present in hops <strong>and</strong> beer; they are only slightly<br />
bitter but display strong foam-stabilizing activity.<br />
Probably the most important oxidation products of the -acids (8.2) (Fig. 8.16) are the<br />
hulupones (8.85) obtained by autoxidation. The hydroperoxide (8.84) has been proposed<br />
as an intermediate but has not been isolated. However, the saturated hydroperoxide,<br />
derived from (8.84), has been isolated from the autoxidation of hexahydrocolupulone.<br />
The hulupones (8.85) are reported to be twice as bitter as the iso- -acids; they are not<br />
found in green hops but accumulate during storage when concentrations of 3% have been<br />
reported. However, they are formed when hops are macerated in air in a blender so the<br />
analytical results are probably high. They are also formed in wort boiling <strong>and</strong> survive into<br />
beer (several ppm). They may be more important in stout brewing when hops are boiled<br />
with wort more than once. After the first boil, the resins that have not dissolved will be<br />
spread over the surface of the spent hops making them accessible to autoxidation between<br />
boils. Probably most claims for better utilization of the -fraction involve oxidation of<br />
the -acids into hulupones. In the laboratory, oxygenation of the -acids in the presence<br />
of sodium sulphite or oxidation with sodium dioxypersulphate, gives better yields of<br />
hulupones. Hulupone (8.85a), pKa 2.6, has been synthesized by the akylation of<br />
dehydrohumulinic acid (8.10a) with 3-methyl-2-butenyl bromide (8.24). With regard to<br />
bitterness of hulupones, Verzele <strong>and</strong> De Keukeleire (1991, p. 377) report that beers<br />
bittered only with pure hulupone (100 mg/l) were undrinkable but beers with 100 IBU<br />
derived from iso- -acids are also likely to be unacceptable.<br />
Autoxidation of hulupones in boiling ethanol gives hulupinic acid (8.86) which lacks<br />
bitterness but has been found in the hard resin of old hops (0.05%). Apart from the<br />
hulupones, autoxidation of the -acids gives complex mixtures of products. Verzele <strong>and</strong><br />
De Keukeleire (1991) devote 86 pages to oxidation products of the -acids <strong>and</strong> describe<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
H 3C<br />
H 3C<br />
C CH<br />
OH<br />
CH 2<br />
(8.69) 2-Methyl-3-buten-2-ol (8.70) 5,5-Dimethyl-(5H )-furan-2-one<br />
HO<br />
(8.71) trans-Humulinone (8.72) tricyclodehydroisohumulone (TCD)<br />
HO<br />
OH HO<br />
O<br />
HO<br />
O<br />
OH<br />
CO·R<br />
(8.73) (8.74)<br />
O<br />
CO<br />
CO<br />
O<br />
OH<br />
OH<br />
O<br />
CO·R<br />
OH<br />
CO·R<br />
(8.75) (8.76)<br />
over 40 products. Probably most of these occur in trace amounts in old hops <strong>and</strong> beers<br />
brewed therefrom but their importance has yet to be determined.<br />
Although used today principally for their bittering properties, hops were originally<br />
used for their preservative value. Simpson (1993b) has studied the effects of the hop bitter<br />
acids on lactic acid bacteria. He found that the hop bitter acids act as mobile carrier<br />
O<br />
HO<br />
O<br />
CO<br />
O<br />
HO<br />
OH<br />
Fig. 8.14 Oxidation products of the -acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
O<br />
O<br />
O<br />
OH<br />
O<br />
OH<br />
CO·R<br />
CO·R<br />
OH<br />
OH<br />
CO·R<br />
OH
O<br />
ionophores <strong>and</strong> inhibit the growth of beer spoilage organisms by dissipating the<br />
transmembrane pH gradient.<br />
8.3 Hop oil<br />
O<br />
O<br />
O<br />
O<br />
O<br />
(8.77) (8.78)<br />
(8.81)<br />
O<br />
OH<br />
CO·R<br />
CO·R<br />
O<br />
(8.83)<br />
CO·R<br />
8.3.1 Introduction<br />
Hops produce up to 3% of essential oil which is responsible for the pleasant hoppy<br />
aroma of beer. It is produced in the lupulin gl<strong>and</strong>s along with the resins, mostly after<br />
resin synthesis is finished. It is largely from the aroma that the grower judges that the<br />
O<br />
HO<br />
OH<br />
O<br />
O<br />
O<br />
(8.79) (8.80)<br />
O<br />
O<br />
O<br />
OH<br />
O<br />
CO·R<br />
OH<br />
O<br />
O<br />
O<br />
O<br />
O<br />
O<br />
O<br />
(8.82)<br />
O<br />
OH<br />
OH<br />
O<br />
OH<br />
CO·R<br />
CO·R<br />
CO·R<br />
Fig. 8.15 Abeo-iso- -acids (Verzele <strong>and</strong> De Keukeleire, 1991).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
OH<br />
CO·R CO·R<br />
O<br />
HO O OH<br />
(8.2) β-Acids (8.84)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O O<br />
Fig. 8.16 Oxidation products of the -acids.<br />
O<br />
CO·R<br />
O O<br />
(8.85) hulupones<br />
HO OH<br />
O O<br />
(8.86) hulupinic acid
hop is ripe <strong>and</strong> ready to pick. The composition of the essential oil depends on genetic<br />
(cultivar) <strong>and</strong> cultural factors but since most commercial hops are picked at an<br />
equivalent degree of ripeness varietal factors will dominate. Nevertheless, in any one<br />
variety, seedless hops will produce more essential oil than seeded hops. For reviews on<br />
hop oil see Sharpe <strong>and</strong> Laws (1981), Stevens (1987), Moir (1994), Deinzer <strong>and</strong> Yang<br />
(1994) <strong>and</strong> Siebert (1994).<br />
By definition, essential oils are volatile in steam, so most essential oil will be lost<br />
when hops are boiled in wort in a copper open to the atmosphere. To add hop aroma to<br />
their beers brewers either add a portion of choice hops towards the end of the boil (most<br />
lager brewers) or add dry hops to the beer either in cask or conditioning tank (premium<br />
ales). The hops used for late <strong>and</strong> dry hopping are chosen for their choice aroma, which<br />
may be transferred directly to the beer. Brewers are usually prepared to pay a premium<br />
for such choice `aroma' hops. The level of -acids is immaterial since the majority will<br />
not be isomerized. The amount of essential oil in a sample of hops is usually measured by<br />
steam distillation, usually with cohobation whereby the oil is retained in a trap <strong>and</strong> the<br />
denser aqueous phase returns to the boiler together with any water-soluble constituents.<br />
Commercial hop oil, prepared in this way, does not have the true aroma of the hops from<br />
which it was prepared. Essential oil constituents are soluble in most organic solvents <strong>and</strong><br />
CO2 so will be present in most solvent extracts of hops. The most volatile constituents<br />
may be removed with the solvent but this is less likely with CO 2 extracts. Hop oils<br />
obtained from CO 2 extracts by molecular distillation (ambient temperature <strong>and</strong> less than<br />
0.001 mm Hg) smell more like the original hops than steam distilled oils. Such oils may<br />
contain water-soluble constituents which would be washed out of steam distilled oils<br />
obtained by cohobation. It is also known that some essential oil constituents are altered<br />
during steam distillation. Before the advent of CO2 extracts some brewers used<br />
commercial steam-distilled oils to dry hop their beers but most people could distinguish<br />
such beers from those dry hopped normally. Today CO2 extracts, or fractions derived<br />
therefrom, are used to impart hop aromas to beers. Brewers are interested in which<br />
constituents of the essential oil influence the flavour of beer <strong>and</strong>, further, whether the<br />
composition of the essential oil can aid the identification of hop cultivars.<br />
The essential oil of hops is a complex mixture of well over 300 compounds, but it can<br />
be separated into two fractions by chromatography on silica gel. The fraction eluted with<br />
light petroleum consists of hydrocarbons while that subsequently eluted with diethyl<br />
ether consists of oxygen-containing compounds such as alcohols, acids, esters, <strong>and</strong><br />
carbonyl compounds. This latter fraction may also contain traces of sulphur-containing<br />
compounds.<br />
The contribution of the individual constituents towards the overall aroma of hop oil<br />
can be assessed by the use of the Flavour Unit (FU) defined as:<br />
‰Concentration of the flavour compoundŠ<br />
Flavour Unit (FU) =<br />
‰Sensory threshold of the flavour compoundŠ<br />
The FU values depend on how, <strong>and</strong> in what medium, the sensory threshold was measured<br />
but if the concentration of a compound does not exceed the sensory threshold (FU < 1) it<br />
will not have a large influence on the overall flavour. Compounds providing 1±2 FU will<br />
be detectable by the assessor <strong>and</strong> compounds providing more than 2 FU are likely to be<br />
dominant flavours. For example, beers contain 10±60 mg/l of iso- -acids, the sensory<br />
detection level of which is 5±6 mg/l, so the iso- -acids will provide 2±12 FU <strong>and</strong> the<br />
bitterness will be perceived by most drinkers (Baxter <strong>and</strong> Hughes, 2001). Individual hop<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
oil constituents will be discussed later but, in general, the sensory detection levels of<br />
sulphur compounds are much lower than those of the corresponding oxygen compounds<br />
which, in turn, are lower than those of the hydrocarbons.<br />
8.3.2 Hydrocarbons<br />
The hydrocarbon fraction may account for 50±80% of the essential oil <strong>and</strong> the major<br />
components, found in most varieties, are the monoterpene myrcene (8.89) <strong>and</strong> the<br />
sesquiterpenes -caryophyllene (8.136) <strong>and</strong> humulene (8.111) which were characterized<br />
by classical means. Farnesene (8.106), which was first isolated from Saaz (Zatec) hops,<br />
was found tobe present in some cultivars but not in others. For example, it was not found<br />
in the oils of Hallertau, Goldings, Fuggle or Cluster hops. From later plant breeding<br />
studies it is thought that the presence of farnesene is asex-linked character controlled by<br />
a single pair of genes with presence dominant to absence. Similarly, on gas<br />
chromatograms of the hydrocarbons of some varieties, several compounds are eluted<br />
after humulene. Two of these were later identified as -(8.108) <strong>and</strong> -selinene (8.110).<br />
The presence or absence of selinene is also controlled by asingle pair of genes but with<br />
incomplete dominance. Traces of many other terpenes <strong>and</strong> sesquiterpenes have been<br />
detected in hop oil.<br />
As the hop ripens, trace of oxygenated compounds of the essential oil appear first,<br />
then the cyclic sesquiterpenes -caryophyllene <strong>and</strong> humulene, <strong>and</strong> finally the<br />
monoterpene myrcene is formed. The percentage of myrcene probably reflects the<br />
ripeness of the cones but the humulene/caryophyllene ratio is usually constant <strong>and</strong> a<br />
varietal characteristic (Table 7.3). The selinene/caryophyllene ratio also appears to be a<br />
varietal characteristic.<br />
The biosynthesis of the terpenoid compounds in the essential oil uses the same<br />
building blocks as required for the isoprenyl side chains of the hop resins (Fig. 8.3).<br />
Dimethylallyl pyrophosphate (8.34) condenses with amolecule of isopentenyl pyrophosphate<br />
(8.33) to give geranyl pyrophosphate (8.88), the parent of the monoterpenes <strong>and</strong><br />
the source of the side chain in the chalcone (8.61) <strong>and</strong> the related flavanone (8.68). With<br />
another molecule of isopentenyl pyrophosphate, geranyl pyrophosphate forms farnesyl<br />
pyrophosphate (8.105), the parent of the sesquiterpenes. Elimination of pyrophosphoric<br />
acid from geranyl pyrophosphate (8.88) gives the major monoterpene myrcene (8.89).<br />
This <strong>and</strong> other monoterpene relationships are shown in Fig. 8.17. Similarly, elimination<br />
of pyrophosphoric acid from farnesyl pyrophosphate (8.105) gives ( -)farnesene (8.106)<br />
(Fig. 8.18). Cyclization of trans, trans-farnesyl pyrophosphate can give two monocations<br />
(8.107) <strong>and</strong> (8.109); 8.107 can lose aproton to give humulene (8.111) while (8.109) can<br />
give -(8.108) <strong>and</strong> -selinene (8.110). Although -caryophyllene (8.136) nearly always<br />
occurs with humulene, it is thought to be formed from atrans, cis-farnesyl cation. An<br />
alternative deprotonation of 8.109 can lead to germacrene B(8.112), germacrene D<br />
(8.113) <strong>and</strong> bicyclogermacrene (8.114). Cyclization of germacrene Bcan give selina-<br />
3,7(11)-diene (8.115) <strong>and</strong> selina-4(15),7(11)-diene (8.116). Germacrene D(Fig. 8.19) is<br />
readily converted into amixture of -(8.117) <strong>and</strong> -muurolene (8.120) <strong>and</strong> -(8.118)<br />
<strong>and</strong> -cadinene(8.121).Allthesesesquiterpenestogetherwithcopaene(8.119)havebeen<br />
identified in hop oil.<br />
Tressl et al. (1993) found at least 15 tricyclic sequiterpenes in the essential oil of the<br />
German variety Hersbrucker spaÈt. These hydrocarbons are probably derived from<br />
bicyclogermacrene (8.114, Fig. 8.20) <strong>and</strong> include viridiflorene (8.126), alloaromadendrene<br />
(8.127), aromadendrene (8.128) <strong>and</strong> -gurjenene (8.129). Traces of the diterpenes<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 8.17 Monoterpene relationships in hop oil.<br />
m- <strong>and</strong> p-camphorene have been found in hop oil but these are thought to be artefacts<br />
formed by a Diels-Alder reaction between two molecules of myrcene. Few, if any, of<br />
these hydrocarbons survive wort boiling but traces may be found in late <strong>and</strong> dry hopped<br />
beers.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(8.99) 8(9)-Menthene (8.100) p-Cymene<br />
(8.101) α-Pinene (8.102) β-Pinene (8.103) camphene<br />
Fig. 8.17 Continued<br />
Guadagni et al. (1966) found the following sensory thresholds: myrcene, 13;<br />
caryophyllene, 64; <strong>and</strong> humulene, 120 ppb. Thus, in the hydrocarbon fraction myrcene is<br />
by far the most potent odorant. In an oil from Brewers' Gold hops, myrcene (63%)<br />
accounted for 58% of the odour units. Similarly, Steinhaus <strong>and</strong> Schieberle (2000) found<br />
that myrcene was the most potent odorant in Spalter Select hops.<br />
8.3.3 Oxygen-containing components<br />
The oxygenated fraction of hop oil is even more complex than the hydrocarbon<br />
fraction but these polar constituents are more likely to survive into beer. Probably the<br />
first oxygenated component of hop oil to be characterized was undecan-2-one (methyl<br />
nonyl ketone, luparone), which is now known to be accompanied by other methyl<br />
ketones. Sharpe <strong>and</strong> Laws (1981) report 60 aldehyes or ketones, 70 esters, 50 alcohols,<br />
25 acids, 30 oxygen heterocyclic compounds <strong>and</strong> 30 sulphur-containing compounds in<br />
hop oil. Among the esters both straight chain <strong>and</strong> branched chain fatty acids <strong>and</strong><br />
alcohols are involved. For example, the methyl esters of hexanoic, octanoic, decanoic,<br />
4-decenoic <strong>and</strong> 4, 8-decadienoic acids are probably by-products of fatty acid<br />
biosynthesis. Branched chain compounds such as 2-methylbutyl isobutyrate presumably<br />
arise from pathways to the carbon skeletons of amino acids. Whether these hop<br />
esters survive into beer is difficult to determine as similar products are formed during<br />
fermentation.<br />
During fermentation methyl 4-decenoate <strong>and</strong> methyl 4, 8-decadienoate undergo<br />
transesterification to produce the corresponding ethyl esters in beer. Probably other esters<br />
behave similarly but brewing yeasts are known to produce esterases. When methyl<br />
heptanoate was added to a fermentation only 35% of the parent acid was recovered with<br />
little methyl or ethyl heptanoate. Nevertheless, transesterification of geranyl pyrophosphate<br />
(8.88) is likely to be the source of geranyl acetate, propionate <strong>and</strong> isobutyrate<br />
which give a floral note to hop oil. Mild (enzymatic) hydrolysis of these esters gives the<br />
primary alcohol geraniol (8.90, R ˆ H) but under acid conditions the tertiary alcohol<br />
linalol (linalool, 8.87, Fig. 8.17) is formed. cis,trans-Isomerization of geraniol gives nerol<br />
(8.93, R ˆ H) <strong>and</strong> neryl esters have been found in hop oil. Oxidation of geraniol <strong>and</strong><br />
nerol gives citral, a mixture of the aldehydes geranial (8.91) <strong>and</strong> neral (8.94) which on<br />
further oxidation give geranic acid (8.92 <strong>and</strong> 8.95), the methyl ester of which occurs in<br />
hop oil.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
OH<br />
O· P · P<br />
(8.104) nerolidol (8.105) farnesyl pyrophosphate<br />
+<br />
+<br />
O· P · P<br />
(8.106) β-Farnesene (8.107) (8.108) α-Selinene<br />
H<br />
(8.109) (8.110) β-Selinene (8.111) humulene<br />
(8.112) germacrene B (8.113) germacrene D (8.114) bicyclogermacrene<br />
H H<br />
(8.115) selina-3,7(11)-diene (8.116) selina-4(15),7(11)-diene<br />
Fig. 8.18 Sesquiterpene relationships in hop oil.<br />
Reduction of the 2, 3-double bond in either geraniol or nerol gives citronellol (8.96).<br />
Nerol readily cyclizes to give -terpineol (8.97) present in hop oil <strong>and</strong> beer. Dehydration<br />
of -terpineol gives limonene (8.98), the major hydrocarbon of citrus oils but also present<br />
in hop oil. Limonene is probably the precursor of the bicyclic monoterpenes such as -<br />
(9.101) <strong>and</strong> -pinene (8.102) <strong>and</strong> camphene (8.103). Limonene can also disproportionate<br />
into p-cymene (8.100) <strong>and</strong> 8(9)-menthene (8.99). Similarly, mild (enzymatic) hydrolysis<br />
of farnesyl pyrophosphate (8.105) can give farnesol while acid hydrolysis gives nerolidol<br />
(8.104); both have been found in hops <strong>and</strong> beer. For these oxygenated compounds<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H
H<br />
H<br />
H<br />
H<br />
(8.113) germacrene D<br />
Guadagni et al. (1966) found the following odour theshold values: undecan-2-one, 7;<br />
linalol, 6; methyl heptanoate, 4; methyl 4-decenoate, 3; methyl 4, 8-decadienoate, 10; 2methylbutyl<br />
isobutyrate, 14; geranyl acetate, 9; geranyl propionate, 10; <strong>and</strong> geranyl<br />
isobutyrate, 13 ppb. Thus, most of these odorants are more potent than myrcene but their<br />
concentration is usually much lower. For example, in the sample of Brewers' Gold oil<br />
mentioned above, methyl 4-decenoate contributed the largest percentage (3.0%) of the<br />
total odour units from the oxygenated fraction.<br />
During hop storage the proportion of hydrocarbons in the essential oil decreases <strong>and</strong><br />
that of the oxygenated components increases. The increase in volatile acids (3 ! 20%)<br />
after three years storage at 0 ëC was mentioned earlier but these acids will be formed by<br />
oxidation of both resins <strong>and</strong> essential oil components. Carbon-carbon double bonds react<br />
with oxygen to form epoxides (oxiranes); in vitro peracids are the usual reagent. These<br />
three-membered rings are readily opened to give diols which may undergo further<br />
reactions. Monoterpene epoxides have not been isolated from hop oils but may be<br />
H<br />
H H<br />
(8.117) α-Muurolene (8.118) δ-Cadinene (8.119) copaene<br />
(8.120) γ-Muurolene (8.121) γ-Cadinene (8.122) α-Cadinene<br />
H<br />
H<br />
OH OH<br />
H<br />
H<br />
H<br />
H<br />
H<br />
H<br />
(8.123) δ-Cadinol (8.124) α-Cadinol (8.125) T-Cadinol<br />
Fig. 8.19 Sesquiterpene relationships: Germacrene D <strong>and</strong> the Cadinenes.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H<br />
OH
H<br />
(8.126)<br />
viridiflorene<br />
H<br />
H<br />
OH<br />
–H +<br />
H<br />
H<br />
+<br />
(8.114) bicyclogermacrene<br />
–H +<br />
O<br />
H<br />
H<br />
H +<br />
–H +<br />
(8.127)<br />
alloaromadendrene<br />
HO<br />
H<br />
intermediates in, for example, the autoxidation of myrcene which leads to linalol,<br />
geraniol, nerol, citral (8.91 <strong>and</strong> 8.94) together with the cyclic limonene. In contrast,<br />
sesquiterpene epoxides are found in hop oil <strong>and</strong> their concentration increases during hop<br />
storage. -Caryophyllene (8.136) (Fig. 8.21) can theoretically form two mono-epoxides<br />
but only one (8.137) has been found in hop oil. Some workers, but not others, have found<br />
H +<br />
H<br />
H<br />
H<br />
H H<br />
(8.128)<br />
aromadendrene<br />
H<br />
H<br />
OH<br />
–H +<br />
(8.129)<br />
α-Gurjunene<br />
(8.130) alloaromadendrene epoxide (8.131) aromadendrene epoxide<br />
(8.132) ledol (8.133) viridiflorol (8.134) globulol (8.135) epiglobulol<br />
Fig. 8.20 Sesquiterpenoids from Bicyclogermacrene.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
+<br />
O<br />
H<br />
H<br />
OH
H H<br />
H O<br />
H H H HO<br />
(8.136) caryophyllene (8.137) caryophyllene epoxide (8.138) caryolan-1-ol<br />
Fig. 8.21 Reactions of Caryophyllene.<br />
it in beer. It undergoes hydrolysis to give at least six products (Deinzer <strong>and</strong> Yang, 1994).<br />
Similarly, caryolan-1-ol (8.138), obtained by acid treatment of caryophyllene, has been<br />
found in beer by some workers but not others.<br />
Humulene (8.111) (Fig. 8.22) can form three mono-epoxides (8.139), (8.140) <strong>and</strong><br />
(8.141). Of these humulene epoxide I(8.139) is most resistant to hydrolysis <strong>and</strong> most<br />
likelytobefoundinbeer.HumuleneepoxideII(8.140)<strong>and</strong>humuleneepoxide III(8.141)<br />
can be interconverted <strong>and</strong> hydrolysis of either gives amixture of at least 30 products;<br />
humulol (8.146) <strong>and</strong> humulenol II (8.147) predominate in beer. Although only traces of<br />
humulene diepoxides are found in fresh hops the amount increases during storage.<br />
Theoreticallytherearesixpairsofenantiomersofhumulenediepoxidebutonlyfivewere<br />
formed by treatment of humulene with m-chloroperbenzoic acid: humulene diepoxides<br />
A-E(8.142±8.145).HumulenediepoxideAisthemostabundantisomer<strong>and</strong>givesatleast<br />
ten products on hydrolysis. There is considerable evidence that sesquiterpene oxidation<br />
products, <strong>and</strong>/or hydrolysis products therefrom, contribute to the hoppy aroma of beer<br />
although the actual compound(s) responsible have not been identified. Further, many<br />
lager brewers regard the aroma produced by the traditional European varieties such as<br />
Hallertauer mittelfruÈh,Hersbrucker<strong>and</strong>Tettnang±theso-called`noble'aroma±superior<br />
to that produced by other varieties. These hops produce high levels of sesquiterpene<br />
oxidation products <strong>and</strong> some brewers store their `aroma' hops for aperiod to facilitate<br />
this oxidation. Humuladieneone (8.148) has been found in beer <strong>and</strong> associated with a<br />
hoppy aroma (Shimazu et al., 1974). It is not acommon oxidation product of humulene<br />
but it may be formed from humulenol II.<br />
The minor sesquiterpenes will react with oxygen in asimilar manner. -Selinene<br />
(8.110, Fig. 8.23) can form two monoepoxides but only one (8.149) has been found in<br />
hop oil together with the related tertiary alcohol, selin-11-en-4-ol (8.151). Reduction of<br />
the other epoxide (8.150) would give -eudesmol (8.153) which with -(8.154) <strong>and</strong> -<br />
eudesmol (9.155) is found in hop oil. In asimilar manner selina-3,7(11)-diene (8.115)<br />
<strong>and</strong> selina-4(15),7(11)-diene (8.116) would give juniper camphor (8.152) found in hop<br />
oil. Analogous reactions in the cadinene series (Fig. 8.19) lead to -(8.124), -(8.123)<br />
<strong>and</strong> T-cadinol (8.125) also found in hop oil <strong>and</strong> beer. The tricyclic sesquiterpenes found<br />
in Hersbrucker spaÈt hops (Fig. 8.20) can also form epoxides. Reduction of<br />
alloaromadendrene epoxide (8.130) gives ledol (8.132) <strong>and</strong> viridiflorol (8.133) while<br />
aromadendrene epoxide (8.131) gives globol (8.134) <strong>and</strong> epiglobol (8.135). Traces of<br />
these compounds probably occur in beers brewed with these hops.<br />
Various oxygen heterocyclic compounds found in hop oil <strong>and</strong> beer are shown in Fig.<br />
8.24. Compounds (8.156±8.160) were first characterized in Japanese hops but later found<br />
in German Spalter hops where the concentration was found to increase on storage. They<br />
are reported to contribute aflowery note to hop aroma. Various furans such as 5,5-<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
H
(8.139)<br />
humuleme epoxide I<br />
O<br />
O<br />
O O<br />
O<br />
(8.142)<br />
humulene diepoxide A<br />
O<br />
(8.111)<br />
humulene<br />
(8.140)<br />
humulene epoxide II<br />
O<br />
O O<br />
O<br />
O OH OH<br />
(8.145)<br />
humulene diepoxide D & E<br />
(8.143)<br />
humulene diepoxide B<br />
(8.146)<br />
humulol<br />
O<br />
(8.148)<br />
humuladienone<br />
Fig. 8.22 Reactions of Humulene.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
(8.141)<br />
humulene epoxide III<br />
(8.144)<br />
humulene diepoxide C<br />
(8.147)<br />
humulenol II
H<br />
H O H<br />
H<br />
(8.110) β-Selinene (8.149) β-Selinene epoxide (8.116) selina-4(15),7(11)-diene<br />
O<br />
HO<br />
H H<br />
HO<br />
(8.150) (8.151) selina-11-en-4-ol (8.152) juniper camphor<br />
H OH H OH<br />
OH<br />
(8.153) β-Eudesmol (8.154) α-Eudesmol (8.155) γ-Eudesmol<br />
Fig. 8.23 Reactions of Selinene.<br />
dimethyl-(5H)-furan-2-one (8.70), 2-hexyl-5-methylfuran, dendrolasin, perillene <strong>and</strong><br />
compoundssuchas(8.163,RˆPr i ,Bu i orCHMeEt)havebeenfoundinhopoil<strong>and</strong>beer<br />
(Moir, 1994). Probably more important for the overall hop aroma are the traces of -<br />
ionone (8.164), -damascenone (8.165) <strong>and</strong> cis-jasmone (8.166) reported. These have<br />
verylowthresholdvalues(seeTable8.6onpage299);itisreportedthatsomeindividuals<br />
can smell as little as 50 fg (10 15 g) of -damascenone.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
O<br />
HO<br />
O<br />
O<br />
O<br />
(8.156) (8.157) (8.158)<br />
O<br />
(8.159) hop ether (8.160) karahana ether (8.161) rose oxide<br />
(8.162) linalol oxide (8.163)<br />
O<br />
O O<br />
Fig. 8.24 Oxygen heterocyclic compounds in hop oil.<br />
8.3.4 Sulphur-containing compounds<br />
Only traces of sulphur-containing compounds are found in hop oil but many have very low<br />
flavour thresholds (Table 8.4). They can be detected by gas chromatography using either a<br />
flame photometric or a Sievers' chemiluminescence detector. Hops in the field may be<br />
treated with elemental sulphur to control mildew <strong>and</strong> it used to be common <strong>practice</strong> to burn<br />
sulphur in the oast. Both of these treatments can influence the spectrum of sulphur-containing<br />
compounds in the oil. For example, the sesquiterpenes caryophyllene <strong>and</strong> humulene can react<br />
with elemental sulphur under mild conditions to give episulphides such as (8.137), (8.139)<br />
<strong>and</strong> (8.140) where sulphur replaces oxygen. The level of these compounds is higher in oils<br />
that have been steam distilled at 100 ëC than in those obtained by vacuum distillation at 25 ëC.<br />
Thus more of these compounds will be introduced into beer by late hopping than by dry<br />
hopping. Myrcene also reacts with sulphur but less readily than the sesquiterpenes. However,<br />
with a suitable activator, a mixture of at least ten sulphur-containing compounds was formed<br />
of which (8.167) is the major component <strong>and</strong> this has been found in hop oil.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
O<br />
O<br />
O<br />
R
Table 8.4 Sulphur compounds in the essential oil of hops<br />
Name Structure Flavour threshold<br />
(ppb)<br />
Methanethiol CH 3SH 0.02<br />
Dimethyl sulphide CH 3.S.CH 3 7.5<br />
Dimethyl disulphide CH 3.S.S.CH 3 7.5<br />
Dimethyl trisulphide CH3.S.S.S.CH3 0.1<br />
(2, 3, 4-Trithiapentane)<br />
2, 3, 5-Trithiahexane CH3.S.S.CH2.S.CH3 ±<br />
Dimethyl tetrasulphide CH3.S.S.S.S.CH3 0.2<br />
(2, 3, 4, 5-Tetrathiahexane)<br />
S-Methyl 2-methylpropanethioate (CH 3) 2 CH.CO.S.CH 3 40 (5)<br />
S-Methyl 2-methylbutanethioate CH 3CH 2CH(CH 3)CO.S.CH 3 1<br />
S-Methyl 3-methylbutanethioate (CH 3) 2CH.CH 2.CO.S.CH 3 50<br />
S-Methyl pentanethioate CH3.(CH2)3.CO.S.CH3 10<br />
S-Methyl 4-methylpentanethioate (CH3)2CH.CH2CH2CO.S.CH3 15<br />
S-Methyl hexanethioate CH3(CH2)4CO.S.CH3 1<br />
S-Methyl heptanethioate CH3(CH2)5CO.S.CH3 ±<br />
S-Methylthiomethyl CH3CH2CH(CH3)CO.S.CH2S.CH3 1<br />
2-Methylbutanethioate<br />
S-Methylthiomethyl (CH 3) 2CH.CH 2CO.S.CH 2S.CH 3 ±<br />
3-Methylbutanethioate<br />
3-Methylthiophene (8.168) 500<br />
3-(4-Methylpent-3-enyl)thiophene (8.169) ±<br />
4-(4-Methylpent-3-enyl)-3, 6- (8.167) 10<br />
dihydro-1, 2-dithiine<br />
4, 5-Epithiocaryophyllene (8.137) with S in place of O 200<br />
1, 2-Epithiohumulene (8.139) with S in place of O 1800<br />
4, 5-Epithiohumulene (8.141) with S in place of O 1500<br />
2-Methyl-5-thiahex-2-ene (CH 3) 2C ˆ CH.CH 2.S.CH 3 0.2<br />
Methylthiohumulene C15H24.S.CH3 ±<br />
Polysulphides also occur in hop oil. Dimethyl trisulphide (2, 3, 4-trithiapentane) is<br />
found only in hops that have not been treated with sulphur (dioxide) on the kiln. It is<br />
formed during steam distillation at 100 ëC from (S)-methylcysteine sulphoxide<br />
(CH3SO.CH2CH(NH2)CO2H) which is destroyed when sulphur is burnt on the kiln but<br />
slowly regenerates during storage. Dimethyl tetrasulphide <strong>and</strong> 2, 3, 5-trithiahexane have<br />
also been found in hop oil. These polysulphides have cooked vegetable, onion-like,<br />
rubbery sulphur aroma with low thresholds.<br />
Thioesters are also present in hop oil (> 1000 ppm), the concentration of which does<br />
not appear to be influenced by treatment of the hops with sulphur or sulphur dioxide.<br />
They appear to be formed by the action of heat so only low levels will be introduced<br />
into beer by dry hopping. Few of the sulphur compounds discussed survive wort<br />
boiling but late addition of hops introduces traces of these compounds, including<br />
thioesters, into wort. During fermentation dimethyl trisulphide <strong>and</strong> some thioesters are<br />
lost but some sulphur volatiles survive into beer, in particular, S-methyl 2methylbutylthioate.<br />
This ester <strong>and</strong> S-methyl hexanethiolate are the major thioesters<br />
introduced into beer by dry hopping. In the sample of Brewers' Gold hops discussed<br />
above Guadagni et al. (1966) found that methyl thiohexanoate contributed 4.8% of the<br />
total odour units.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
8.3.5 Most potent odorants in hop oil<br />
The pioneering work of Guadagni et al. (1966), discussed above, suggests that<br />
myrcene, methyl thiohexanoate, methyl 4-decenoate, caryophyllene <strong>and</strong> humulene are<br />
the most potent odourants in Brewers' Gold hops. Using a gas chromatographyolfactometry<br />
(GC-O) technique called `Osme' Sanchez et al. (1992) found that linalol<br />
(8.87), neral (8.94) <strong>and</strong> humulene epoxide III (8.141) were the most potent odorants in<br />
Hallertau mittelfruÈh <strong>and</strong> Mount Hood hops. Steinhaus <strong>and</strong> Schieberle (2000) used<br />
another GC-O technique ± Aroma Extract Dilution Analysis (AEDA) ± to examine an<br />
extract of Spalter Select hops. First, 36 areas were identified on the gas chromatogram<br />
by olfactory analysis. The extract was then diluted with an equal volume of diethyl<br />
ether <strong>and</strong> the analysis repeated. Further dilutions were made until no aroma was<br />
detected. The Flavour Dilution value (FD) is the highest dilution at which the aroma<br />
can be detected. The results of this study with dried <strong>and</strong> undried Spalter Select hops is<br />
given in Table 8.5 together with the results of a headspace analysis of the dried hops.<br />
Twenty-three odorants (FD > 4) were identified. In green undried hops (Z)-3-hexenal<br />
<strong>and</strong> linalol were the most potent odorants. Much of the (Z)-3-hexenal was lost during<br />
drying after which the most potent odorant was trans-4, 5-epoxy-(E)-2-decenal (8.171),<br />
probably formed by thermal cleavage of 12, 13-epoxy-9-hydroperoxy-11-octadecenoates<br />
(8.170).<br />
Table 8.5 Most odour-active compounds in hops cv. Spalter Select (Flavour Dilution values)<br />
(Steinhaus <strong>and</strong> Schieberle, 2000)<br />
Undried Dried Headspace*<br />
1a. Ethyl 2-methylpropionate 128 128 32<br />
2. Methyl 2-methylpropionate 256 128 16<br />
3a. (Z)-3-hexenal 2048 16<br />
3b. Hexanal 16<br />
4. Ethyl 2-methylbutanoate 32 16 16<br />
9. Propyl 2-methylbutanoate 16 64 8<br />
Dimethyl trisulphide 16<br />
11. 1-Octen-3-one 32 32 1<br />
12. (Z)-1, 5-Octadien-3-one 32 32 2<br />
13. Myrcene 512 1024 256<br />
Octanal 8 8 1<br />
17. Phenylacetaldehyde 16<br />
20. Linalol 2048 2048 256<br />
21. Nonanal 64 218<br />
14. (E, Z)-2, 6-Nonadienal 16 4 1<br />
23a. 1, 3(E), 5(Z)-Undecatriene 128<br />
23b. 1, 3(E), 5(Z),9-Undecatetrene 128<br />
15. 4-Ethenyl-2-methoxyphenol 32 32<br />
trans-4, 5-Epoxy-(E)-2-decenal 512 4096<br />
33. Humulene 16 8<br />
37. Butanoic acid 32<br />
38. 2-Methylbutanoic acid 64<br />
3-Methylbutanoic acid<br />
39. Pentanoic acid 4<br />
* Relative FD. Unidentified compounds (FD < 32) omitted.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH3·(CH2)4·CH<br />
O<br />
CH3·(CH2)4·CH<br />
C CH·CH2·CH·(CH2)7·CO2R<br />
O<br />
Δ<br />
OOH<br />
(8.170) 12,13-epoxy-9-hydroperoxy-11-octadecenoate<br />
CH·CH CH·CHO<br />
(8.171) trans-4,5-epoxy-(E)-2-decenal<br />
After this fatty acid oxidation product, linalol <strong>and</strong> myrcene were the most potent<br />
odorants in dried hops followed by ethyl <strong>and</strong> methyl 2-methylpropionate, (Z)-1,5octadien-3-one,<br />
nonanal, 1,3(E),5(Z)-undecatriene <strong>and</strong> 1,3(E),5(Z),9-undecatetraene. In<br />
the headspace of the dried hops myrcene <strong>and</strong> linalol were again the most potent odorants.<br />
The sesquiterpene epoxides are probably not sufficiently volatile to be found in these<br />
studies. It is noteworthy that none of the odorants characterized in these studies was<br />
classed as `hoppy'. This suggests that the typical hop aroma is probably due to a<br />
synergistic effect. However, it is possible there is an, as yet unknown, highly potent<br />
`hoppy' odorant (Siebert, 1994).<br />
8.3.6 Hop oil constituents in beer<br />
Quantitatively, any contribution that the hop makes to beer volatiles will be dwarfed by<br />
ethanol <strong>and</strong> other volatile products of fermentation. Also, during fermentation, carbonyl<br />
compounds<strong>and</strong>possiblyepoxidesmaybereducedtoalcohols.Thefirstdetailedstudyof<br />
hop volatiles in aGerman Pilsener beer was by Tressl et al. (1978) who characterized 47<br />
compounds in beer which were known constituents of Spalter hops. These included 28<br />
terpenoids <strong>and</strong>sesquiterpenoids (Table 8.6). Also included in Table 8.6 are the threshold<br />
values, collected from various sources, from which it is possible to judge which<br />
compounds make an important contribution to the overall flavour. However, it is often<br />
difficult to obtain some of these odorants sufficiently pure to obtain accurate threshold<br />
values (Siebert, 1994). From Table 8.6 it appears that hop ether (8.159), karahana ether<br />
(8.160), linalol <strong>and</strong>, perhaps, -ionone <strong>and</strong> -damascenone make important contributions<br />
to the hop aroma in beer. Humulene epoxide I may contribute but humulene<br />
epoxide II, humulol <strong>and</strong> humuleneol II do not produce sufficient flavour units. However,<br />
the hydrolysis mixture from humulene epoxide II, with a cedar, lime, banana character,<br />
may influence the overall flavour (Deinzer <strong>and</strong> Yang, 1994). Goiris et al. (2002)<br />
obtained an oxygenated sesquiterpene fraction which when added to beer after<br />
fermentation produced spicy or herbal flavour notes reminiscent of typical `noble' hop<br />
aroma.<br />
Peacock et al. (1980) examined pilot brews made with Hallertau mittelfruÈh <strong>and</strong><br />
Washington Cluster hops <strong>and</strong> found that the Hallertau brew contained higher levels of -<br />
terpineol, humulene epoxide I, humulol, T-cadinol (8.125), -eudesmol (8.154) <strong>and</strong><br />
humulenol II than the Cluster brew. Similarly, in an American commercial beer, brewed<br />
mainly with Hallertau mittelfruÈh hops, caryolan-1-ol (8.138), humulene epoxide I, -<br />
cadinol (8.123) <strong>and</strong> -eudesmol were found but these compounds were not detected in<br />
beers brewed with Cascade hops. Indeed Cascade hops introduced a floral hop aroma into<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 8.6 Terpenoids <strong>and</strong> sesquiterpenoids characterized in beer (Tressl et al., 1978)<br />
Compound Approx. conc. (ppb) Threshold value (ppb)<br />
(8.153) 5 ±<br />
(8.152) 10 ±<br />
(8.151) 50 ±<br />
Hop ether (8.159) 35 5 (in water)<br />
Karahana ether (8.160) 60 5 (in water)<br />
trans-Linalol oxide (8.157) 20 ±<br />
Humuladienone (8.148) 10 100<br />
Caryophyllene epoxide (8.139) 18 ±<br />
Humulene epoxide I (8.137) 125 10 (in water)<br />
Humulene epoxide II (8.140) 40 450<br />
Linalol (8.87) 470 27<br />
-Fenchyl alcohol 40 ±<br />
Terpinen-4-ol 15<br />
-Terpineol (8.97) 40 2000<br />
Citronellol (8.96) 10 ±<br />
Geraniol (8.90) 5 36<br />
Caryolan-1-ol (8.138) 25 ±<br />
Nerolidol (8.104) 25 ±<br />
Juneol 5 ±<br />
Epi-Cubenol 20 ±<br />
Caryophyllenol 5 ±<br />
T-Cadinol (8.125) 45 ±<br />
Humulol (8.146) 220 2000<br />
-Cadinol (8.123) 35 ±<br />
Humulenol II (8.147) 1150 2500<br />
-Ionone (8.164) 1 0.007<br />
-Damascenone (8.165) Tr 0.002<br />
cis-Jasmone (8.166) 10 ±<br />
All the above compounds were found in Spalter hops.<br />
beer (Peacock et al., 1981). Cascade hops contain high levels of linalol, geraniol <strong>and</strong><br />
geranyl isobutyrate <strong>and</strong> produce beers in which the level of linalol <strong>and</strong>, in particular,<br />
geraniol exceed the threshold value (the threshold value of geranyl isobutyrate, 450 ppb,<br />
was not exceeded). All the hops examined contained linalol <strong>and</strong> Clusters, Talisman <strong>and</strong><br />
Shin-shu-wase also contained high levels of geraniol. Geraniol was not found in<br />
Hallertau, Hersbrucker <strong>and</strong> Perle hops.<br />
Moir et al. (1983) compared the hop oil constituents in copper hopped, late hopped<br />
<strong>and</strong> dry hopped beers. They found 40±75ppb of hop oil constituents in copper hopped<br />
beer, the major constituents being humulene-8,9-epoxide (8.139) <strong>and</strong> -humulen-1-ol.<br />
In late hopped lagers they found, in addition, linalol, methyl geranate, 2-nonanol <strong>and</strong> 2undecanol.<br />
In dry hopped ales they found myrcene, linalol, 2-undecanone, <strong>and</strong> the<br />
esters isobutyl isobutyrate, isoamyl isobutyrate, isoamyl isovalerate <strong>and</strong> methyl 4decenoate.<br />
Nickerson<strong>and</strong>VanEngel(1992),followingFoster<strong>and</strong>Nickerson(1985),regardedthe<br />
22 components listed in Table 8.7 as being important for the hop aroma of beer. It will be<br />
noted that in addition to the oxidation products <strong>and</strong> the floral-estery compounds they<br />
included anumber of citrus-piney compounds. They measured the concentration of these<br />
compounds individually <strong>and</strong> as agroup in hops (nl/g) <strong>and</strong> in beer ( l/l). They regarded<br />
thesumoftheamountofoilcomponentsinthethreegroupsasHopAromaUnits.Having<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 8.7 Classification of hop aroma constituents (Nickerson <strong>and</strong> Van Engel, 1992)<br />
Oxidation products Floral compounds Citrus-piney compounds<br />
(Caryolan-1-ol) Geraniol -Cadinene<br />
Caryophyllene oxide Geranyl acetate -Cadinene<br />
(Humulene diepoxide A) Geranyl isobutyrate (Citral)<br />
Humulene diepoxide B Linalol Limonene<br />
(Humulene diepoxide C) (Limonene-10-ol)<br />
Humulene epoxide I -Muurolene<br />
Humulene epoxide II (Nerol)<br />
Humulene epoxide III -Selinene<br />
Humuleneol II<br />
Humulol<br />
Compounds not usually detected in steam-distilled oil from fresh hops are in parentheses.<br />
determined the Hop Aroma Units needed to produce a desirable beer, Van Engel <strong>and</strong><br />
Nickerson (1992) used this concept to calculate the late hop addition required to produce<br />
beers of consistent hoppiness. They found that during hop storage the Aroma Units <strong>and</strong><br />
the amount of - <strong>and</strong> -acids varied independantly of each other. Therefore, the<br />
bitterness was determined by the addition of bitter hops to the kettle after making<br />
allowance for the -acids in the aroma hops.<br />
Lermusieau et al. (2001) used aroma extract dilution analysis (AEDA) to compare<br />
three beers; one brewed without hops, one brewed with Saaz hops <strong>and</strong> one brewed<br />
with Challenger hops (added seven minutes before the end of the 75 min. boil). Fortyfive<br />
odours were detected in the unhopped beer with an additional 15 in the beers<br />
treated with Saaz hops <strong>and</strong> 16 in the beers treated with Challenger hops. Those with an<br />
FD >16 in the hopped beers included dimethyl trisulphide, -nonalactone <strong>and</strong><br />
components suspected to be N-methylmercaptoacetamide, -damascenone <strong>and</strong> ethyl<br />
cinnamate. An unknown compound in hopped beers (initially thought to be 3mercaptobutan-2-ol)<br />
gave the highest FD values <strong>and</strong> was also present in the hops used.<br />
Saaz pellets were readily distinguishable from Challenger pellets by AEDA;<br />
Challenger pellets were richer in sulphur compounds notably dimethyl disulphide<br />
<strong>and</strong> diethyl disulphide.<br />
8.3.7 Post fermentation aroma products<br />
Products added to bright beer must be completely soluble to achieve 100% utilization<br />
<strong>and</strong> a reproducible flavour. Further filtration would remove both bitter <strong>and</strong> aroma<br />
products indiscriminately. Addition of whole hop oils <strong>and</strong> products containing them<br />
such as emulsions or oil rich fractions are likely to produce a haze. In hop oil, whether<br />
produced by steam distillation or molecular distillation of a CO2 extract, the<br />
hydrocarbon fraction is much less soluble, <strong>and</strong> contains less FU/g, than the<br />
oxygenated fraction. Thus, starting with a molecularly distilled CO2 extract, which<br />
retains the aroma of the parent hop, by liquid/liquid separation a sesquiterpene-less oil<br />
can be obtained with the true aroma of the hop. A 1% solution of this sesquiterpeneless<br />
oil in ethanol is known as a Dry Hop Essence which can be added to bright beer<br />
(250 g/l) to impart a hoppy aroma. Such dry hop essences can be prepared from a<br />
single cultivar of hops giving more control of the hop aroma produced (Marriott,<br />
1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 8.8 Flavour descriptors associated with late hop essences (Marriott, 1999)<br />
Spicy-LHE Floral-LHE Estery-LHE Citrussy-LHE<br />
S<strong>and</strong>alwood Geranium Pear skin Lemon<br />
Oakmoss Neroli Pineapple Grapefruit<br />
Astringent Rose Fruity Bergamot<br />
Synergist Lavender Sweet Lime<br />
It is generally held that late hopping <strong>and</strong> dry hopping produce different aromas in<br />
beers; late-hopped beers are said to have hoppy, fruity-citrus, floral <strong>and</strong> fragrant<br />
flavours. By column chromatography of aCO2 extract Westwood <strong>and</strong> Daoud (1985)<br />
produced aLate Hop Essence, which when added (150±200ppb) to bright beer could<br />
not be distinguished from alate-hopped beer. This essence contained at least 130<br />
compounds including linalol (6±20%), humulenol, linalol oxide, caryophyllene oxide,<br />
hop ether <strong>and</strong> humulene epoxide(s). Further fractionation (Gardner, 1994) led to four<br />
late hop essences (Table 8.8). The `spicy' fraction is composed of mono- <strong>and</strong> sesquiterpene<br />
alcohols, the `floral' fraction is principally ketones, epoxides <strong>and</strong> esters, the<br />
`estery' fraction is composed exclusively of methyl esters of C6 to C10 branched <strong>and</strong><br />
straight chain fatty acids, <strong>and</strong> the `citrussy' fraction is acomplex mixture of terpene<br />
alcohols, ketones, <strong>and</strong> C 5to C 8aliphatic alcohols (Marriott, 1999). These essences<br />
are typically used at 50±100 g/l <strong>and</strong> are supplied as 1% solutions in ethanol.<br />
8.4 Hop polyphenols (tannins)<br />
Hops contain 0.8±1.5% polyphenols; the amount <strong>and</strong> composition of this fraction<br />
depends on the cultivar <strong>and</strong> growing history of the hop examined. By HPLC Forster<br />
et al., (1995, 1996) found over 100 compounds in the polyphenol fraction of hops.<br />
Many of these constituents are also found in malt including the hydroxycinnamic<br />
acids, the hydroxybenzoic acids <strong>and</strong> chlorogenic acid (4.133); gallic acid (4.125) has<br />
long been known as aconstituent of hops. Jerumanis (1985) found 817±2821mg/kg<br />
of (+)-catechin (4.138) in hops <strong>and</strong> the related epi-catechin (4.139) has been reported.<br />
Jerumanis also found 428±1472mg/kg of procyanidin B-3 (4.143) <strong>and</strong> 287±875mg/<br />
kg of procyanidin C-2 (8.172) in hops; the highest concentration being found in Saaz<br />
hops. Acid treatment of these procyanidins will liberate the red pigment cyanidin<br />
(4.136) but similar treatment of hop polyphenols also liberates delphinidin (4.135)<br />
indicating that prodelphinidin compounds are present. Indeed the cyanidin:<br />
delphinidin ratio (1.2±6.2) may be avarietal characteristic (McMurrough, 1981). In<br />
beer proanthocyanidins will slowly react with the proteins present to form anonbiological<br />
haze which will limit the shelf-life of bottled beers. Brewers use various<br />
treatments to remove proanthocyanidins <strong>and</strong> increase the shelf-life of their beers<br />
(Chapter 15). Normally, malt contributes more of these compounds to beer than hops<br />
but the breeding <strong>and</strong> use of proanthocyanidin-free barley varieties such as Caminant<br />
<strong>and</strong> Clarity leave hops as the major source of these materials. Beers brewed with a<br />
proanthocyanidin-free malt <strong>and</strong> bittered with apure hop resin extract are said to lack<br />
character.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO<br />
HO<br />
HO<br />
HO<br />
HO<br />
OH<br />
OH<br />
OH<br />
OH<br />
O<br />
OH<br />
OH<br />
OH<br />
HO<br />
(8.172) procyanidin C3 (trim α C8-4, βC8-4, γC)<br />
HO<br />
O<br />
HO<br />
O<br />
O<br />
O·R<br />
(8.173)<br />
It was mentioned above that the lupulin gl<strong>and</strong>s appear to lack the enzyme chalcone<br />
cyclase but this must be present elsewhere in the hop since the flavanols kaempferol<br />
(8.57) <strong>and</strong> quercetin (8.58) are found in the polyphenol fraction mainly as their<br />
glycosides. Fourteen glycosides have been characterized in hops including quercitrin<br />
(8.58, Rˆrhamnosyl, R' ˆH), isoquercitrin (8.58, Rˆ -D-glucosyl, R' ˆH), rutin<br />
(8.58, Rˆ -L-rhamnosido-6- -D-glucosyl) <strong>and</strong> astragalin (8.173, Rˆ -D-glucosyl).<br />
Recently the 3-O- (6'-O-malonylglucosides) of kaempferol <strong>and</strong> quercetin have also been<br />
identified. It is noteworthy that phloroglucinol (8.18) <strong>and</strong> phlorisobutyrophenone -Dglucoside,<br />
putativeintermediatesinthebiosynthesisofthehopresins, have beenfound in<br />
the hop polyphenols.<br />
8.5 Chemical identification of hop cultivars<br />
Hop cultivars can now be determined by DNA typing but this technique is not widely<br />
available so interest remains in identification by chemical means. From the resins the<br />
percentage of cohumulone in the -acids (<strong>and</strong> colupulone in the -acids) is avarietal<br />
factor (Table 7.3) but more information can be obtained from essential oil analyses. By<br />
inspection of the gas chromatograms of the essential oil, or oils obtained by molecular<br />
distillation of CO2 extracts, experienced analysts can usually suggest the cultivar from<br />
which the oil came. Numerically the humulene/caryophyllene ratio (Table 7.3) <strong>and</strong><br />
perhaps the selinene/caryophyllene ratio are varietal characteristics. Stenroos <strong>and</strong> Siebert<br />
(1984) applied pattern recognition <strong>and</strong> multivariate analysis techniques to hop oil<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
O<br />
OH<br />
OH<br />
OH
chromatograms <strong>and</strong> found that SIMCA <strong>and</strong> Stepwise Discriminant Analysis were most<br />
successful in classifying the major hop varieties. Perpete et al. (1998) examined the oils<br />
obtained from pellets of twelve varieties <strong>and</strong> developed a flow chart to distinguish<br />
between them <strong>and</strong> Lermusieau <strong>and</strong> Collin (2001) extended the work to aged samples. De<br />
Cooman et al. (1998) studied only three varieties : Saaz, Nugget <strong>and</strong> Wye Target, but by<br />
three different techniques. The cohumulone ratio only distinguished Saaz from the other<br />
two varieties but essential oil analyses separated all three varieties. They also used<br />
flavonoid analyses, the glycosides of quercetin <strong>and</strong> kaempferol, which readily<br />
distinguished Wye Target from the others. The application of these techniques to more<br />
varieties is promised.<br />
8.6 References<br />
AITKEN, R. A., BRUCE, A., HARRIS, J. O. <strong>and</strong> SEATON, J. C. (1970) J. Inst. <strong>Brewing</strong>, 76, 29.<br />
ALDERTON, G., BAILEY, G. F., LEWIS, J. C. <strong>and</strong> STITT, F. (1954) Analyt. Chem., 26, 983.<br />
ALDERWEIRELDT, F., ANTEUNIS, M., DIERCKENS, J. <strong>and</strong> VERZELE, M. (1965) Bull. Soc. Chim. Belg., 74, 29.<br />
ASHURST, P. R. <strong>and</strong> LAWS, D. (1966) J. Chem. Soc. C., 1615.<br />
ASHURST, P. R. <strong>and</strong> LAWS, D. (1967) J. Inst. <strong>Brewing</strong>, 73, 535.<br />
BAKER, C. W. (1990) Tech. Quart. MBAA, 27, 33.<br />
BAXTER, E. D. <strong>and</strong> HUGHES, P. S. (2001) Beer: Quality, Safety <strong>and</strong> Nutritional Aspects. Royal Society of<br />
Chemistry, Cambridge, pp. xiv + 138.<br />
BENITEZ, J. L., FOSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. <strong>and</strong> WESTWOOD, K.<br />
T. (1997) European Brewery Convention Manual of Good Practice, Hops <strong>and</strong> Hop Products,<br />
pp. xiv + 186. Verlag Hans Carl, NuÈrnberg.<br />
BURTON, J. S., STEVENS, R. <strong>and</strong> ELVIDGE, J. A. (1964) J. Inst. <strong>Brewing</strong>, 70, 345.<br />
CAHN, R. S., INGOLD, C. <strong>and</strong> PRELOG, V. (1951) J. Chem. Soc., 612.<br />
CAHN, R. S., INGOLD, C. <strong>and</strong> PRELOG, V. (1956) Experientia, 12, 81.<br />
CAHN, R. S., INGOLD, C. <strong>and</strong> PRELOG, V. (1966) Angew<strong>and</strong>te Chemie, 78, 413.<br />
CLARKE, B. J. <strong>and</strong> HILDEBRAND, R. F. (1965) J. Inst. <strong>Brewing</strong>, 71, 26.<br />
COLLINS, E. <strong>and</strong> SHANNON, P. V. R. (1973) J. Chem. Soc. Perkin I, 419.<br />
COLLINS, M. <strong>and</strong> LAWS, D. R. J. (1973) J. Chem. Soc. Perkin I, 2013.<br />
COLLINS, M., LAWS, D. R. J., MCGUINESS, J. D. <strong>and</strong> ELVIDGE, J. A. (1971) J. Chem. Soc., 3814.<br />
DE COOMAN, L., EVERAERT, E. <strong>and</strong> DE KEUKELEIRE, D. (1998) Phytochem. Anal., 9, 145.<br />
DE COOMAN, L., AERTS, G., OVERMEIRE, H. <strong>and</strong> DE KEUKLEIRE, D. (2000) J. Inst. <strong>Brewing</strong>, 106, 169.<br />
DEINZER, M. <strong>and</strong> YANG, X. (1994) EBC Monograph XXII Symposium on Hops, Zoeterwoude, p. 181.<br />
DE KEUKELEIRE, D. <strong>and</strong> VERZELE, M. (1970) Tetrahedron, 26, 385.<br />
DE KEUKELEIRE, D. <strong>and</strong> VERZELE, M. (1971) Tetrahedron, 27, 4939.<br />
EISENREICH, W., SCHWARZ, M. CARTAYRADE, A., ARIGONI, D., ZENK, M. H. <strong>and</strong> BACHER, A. (1998)<br />
Chemistry & Biology, 5, R221.<br />
EUROPEAN BREWERY CONVENTION. (1994) Monograph XXII. Symposium on Hops, Zoeterwoude, xviii +<br />
300 pp.<br />
EUROPEAN BREWERY CONVENTION (1997) see Benitez et al.<br />
FERNANDEZ, C. M. (1967) Chem. Commun., 1212<br />
FORSTER, A., BECK, B. <strong>and</strong> SCHMIDT, R. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, p. 143<br />
FORSTER, A., BECK, B. <strong>and</strong> SCHMIDT, R. (1996) Proc. 24th Conv. Asia-Pacific Section Institute of <strong>Brewing</strong>,<br />
Singapore, p. 243.<br />
FOSTER, R. T. <strong>and</strong> NICKERSON, G. B. (1985) J. Amer. Soc. Brew. Chem., 43, 127.<br />
FUNG, S. Y., ZUURBIER, K. W. M., PANIEGO, N. B., SCHEFFER, J. J. C. <strong>and</strong> VERPORTE, R. (1997) Proc. 26th<br />
Congr. Eur. Brew. Conv. Maastricht, 1997, 215.<br />
GARDNER, D. S. J. (1994) in EBC Symposium on Hops, Zoeterwoude, p. 114.<br />
GLASER, G. (2002) Modern Brewery Age, 53 (12), 6, 32.<br />
GOESE, M., KAMMHUBER, K., BACHER, A., ZENK, M. H. <strong>and</strong> EISENREICH, W. (1999) Eur. J. Biochem., 263,<br />
447.<br />
GOIRIS, K., DE RIDDER, M., DE ROUCK, G., BOEYKENS, A., VAN OPSTALE, F., AERTS, G., DE COOMAN, L. <strong>and</strong><br />
DE KEUKELEIRE, D. (2002) J. Inst Brew., 108, 86.<br />
GOLDSTEIN, H., TING, P., NAVARRO, A. <strong>and</strong> RYDER, P. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes,<br />
p. 53.<br />
GREEN, C. P. (1978) J. Inst. <strong>Brewing</strong>, 84, 312.<br />
GUADAGNI, D. G., BUTTERY, R. G. <strong>and</strong> HARRIS, J. (1966) J. Sci. Food Agric., 17, 142.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HAY, B. A. <strong>and</strong> HOMISKI, J. W. (1991) J. Agric. Food Chem., 39, 1732.<br />
HERMANS-LOCKKERBOL, A. C. J. <strong>and</strong> VERPORTE, R. (1994) J. Chromatography A, 664, 45.<br />
HEYERICK, A., ZHAO, Y., SANDRA, P., HUVAERE, K., ROELENS, F. <strong>and</strong> DE KEUKELEIRE, D. (2003) Photochem.<br />
Photobiol. Sci., 2, 306.<br />
HUGHES, P. S. (1996) J. Chromatography A, 731, 327.<br />
HUGHES, P. S. (2000) J. Inst. <strong>Brewing</strong>, 106, 271.<br />
HUGHES, P. S. <strong>and</strong> SIMPSON, W. J. (1996) J. Amer. Soc. Brew. Chem., 54, 234.<br />
HUGHES, P. S., MENNNEER, I. D., WALTERS, M. T. <strong>and</strong> MARINOVA, G. (1997) Proc. 26th Congr. Eur. Brew.<br />
Conv., Maastricht, p. 231.<br />
JERUMANIS, J. (1985) J. Inst. <strong>Brewing</strong>, 91, 250.<br />
KANEDA, H., KANO, Y., KOSHINO, S. <strong>and</strong> OHYA-NISHIGUCHI (1992) J. Agric. Food Chem., 40, 2102.<br />
KOLLER, H. (1969) J. Inst. <strong>Brewing</strong>, 75, 175.<br />
LERMUSIEAU, G. <strong>and</strong> COLLIN, S. (2001) J. Am. Soc. Brew. Chem., 59, 39.<br />
LERMUSIEAU, G., BULENS, M. <strong>and</strong> COLLIN, S. (2001) J. Agric. Food Chem., 49, 3867.<br />
LIKENS, S. T., NICKERSEN, G. B. <strong>and</strong> ZIMMERMANN, C. E. (1970) Proc. Am. Soc. Brew. Chem., p. 68.<br />
MACWILLIAM, I. C. (1953) J. Inst. <strong>Brewing</strong>, 59, 142, 480.<br />
MARRIOTT, R. (1999) Proc. 7th Conv. Inst. <strong>Brewing</strong>, Africa Section, Nairobi, p. 140.<br />
MAYE, J. P., MULQUEEN, S., WEIS, S. XU, J. <strong>and</strong> PRIEST, M. (1999) J. Amer. Soc. Brew. Chem., 57, 55.<br />
MCMURROUGH, I. (1981) J. Chromatography, 218, 683.<br />
MOIR, M. (1994) EBC Monograph XXII ± Symposium on Hops, Zoeterwoude, p. 165.<br />
MOIR, M. (2000) J. Amer. Soc. Brew. Chem., 58, 131.<br />
MOIR, M. <strong>and</strong> SMITH, R. J. (1995) Proc. 20th Congr. Eur. Brew. Conv., Brussels, p. 125.<br />
MOIR, M., SEATON, J. C. <strong>and</strong> SUGGERT, A. (1983) Proc. 19th Congr. Eur. Brew. Conv. London, p. 63.<br />
MOLTKE, A. B. <strong>and</strong> MEILGAARD, M. (1955) Brygmesteren, 12, 65.<br />
NICKERSON, G. B. <strong>and</strong> LIKENS, S. T. (1979) J. Amer. Soc. Brew. Chem., 37, 184.<br />
NICKERSON, G. B. <strong>and</strong> VAN ENGEL, E. L. (1992) J. Amer. Soc. Brew. Chem., 50, 77.<br />
NOMENCLATURE SUB-COMMITTEE: HOPS LIAISON COMMITTEE (1969) J. Inst. <strong>Brewing</strong>, 75, 340.<br />
PEACOCK, V. (1998). Tech. Quart. MBAA, 35 (1), 4.<br />
PEACOCK, V. E., DEINZER, M. L., MCGILL, L. A. <strong>and</strong> WROLSTAD, D. E. (1980) J. Agric. Food Chem., 28, 774.<br />
PEACOCK, V. E., DEINZER, M. L., LIKENS, S. T., NICKERSON, G. B. <strong>and</strong> MCGILL, L. A. (1981) J. Agric. Food<br />
Chem., 29, 1265.<br />
PERPEÁ TE, P., MEÂ LOTTE, L., DUPIRE, S. <strong>and</strong> COLLINS, S. (1998) J. Am. Soc. Brew. Chem., 56, 104.<br />
PFENNINGER, H. B., SCHUR, F., VATERLAUS, B. P., SIGG, T. <strong>and</strong> WILD, J. (1975) Proc. 15th Congr. Eur. Brew.<br />
Conv., Nice, p. 159.<br />
POWER, F. B., TUTIN, F. <strong>and</strong> ROGERSON, H. (1913) J. Chem. Soc., 103, 1267.<br />
RIEDL, W. (1951) Brauwissenschaft, 81.<br />
RIEDL, W. (1954) Ann., 585, 38.<br />
RIEDL, W., NICKL, J., RISSE, K. H. <strong>and</strong> MITTELDORF, R. (1956) Ber., 89, 1849.<br />
RIGBY, F. L. (1972) Proc. Ann. Meet. Amer. Soc. Brew. Chem., 46.<br />
RIGBY, F. L. <strong>and</strong> BETHUNE, J. L. (1955) J. Inst. <strong>Brewing</strong>, 61, 325.<br />
RILLAERS, G. <strong>and</strong> VERZELE, M. (1962) Bull. Soc. Chim. Belg., 71, 438.<br />
ROHMER, M. (1998) Progress in Drug Research, 50, 136±154.<br />
SANCHEZ, N. B., LEDERER, C. L., NICKERSON, G. B., LIBBY, M. L. <strong>and</strong> MCDANIEL, M. R. (1992) Food Science<br />
<strong>and</strong> Human Nutrition Charambous, G. (ed.) Elsevier, Amsterdam, p. 370.<br />
SHARPE, F. R. <strong>and</strong> LAWS, D. R. J. (1981) J. Inst. <strong>Brewing</strong>, 87, 96.<br />
SHARPE, F. R. <strong>and</strong> ORMROD, H. (1991) J. Inst. <strong>Brewing</strong>, 97, 31.<br />
SHIMAZU, T., HASHIMOTO, N. <strong>and</strong> KUROIWA, Y. (1974). Proc. Amer. Soc. Brew. Chem., 33, 7.<br />
SIEBERT, K. J. (1994) EBC Monograph XXII. Symposium on Hops, Zoeterwoude pp. 198, 215.<br />
SIGG-GRUTTER, T. <strong>and</strong> WILD, J. (1974) Swiss Patent, 1458 343.<br />
SIMPSON, W. J. (1993a) J. Inst. <strong>Brewing</strong>, 99, 317.<br />
SIMPSON, W. J. (1993b) J. Inst. <strong>Brewing</strong>, 99, 405.<br />
SIMPSON, W. J. <strong>and</strong> SMITH, A. R. W. (1992) J. Applied Bacteriology, 72, 327.<br />
SMITH, R. J., DAVIDSON, D. <strong>and</strong> WILSON, R. J. H. (1998) J. Amer. Soc. Brew. Chem., 56, 52.<br />
SPETSIG, L. O. (1955) Acta Chem. Sc<strong>and</strong>., 9, 1421.<br />
SPETSIG, L. O. (1964) J. Inst. <strong>Brewing</strong>, 70, 440.<br />
STEINHAUS, M. <strong>and</strong> SCHIEBERLE, P. (2000) J. Agric. Food Chem., 48, 1776.<br />
STENROOS, L. E. <strong>and</strong> SIEBERT, K. J. (1984) J. Am. Soc. Brew. Chem., 42, 54.<br />
STEVENS, J. F., IVANCIC, M., HSU, V. L. <strong>and</strong> DEINZER, M. (1997) Phytochemistry, 44, 1575.<br />
STEVENS, J. F., MIRANDA, C. L., BUHLER, D. R. <strong>and</strong> DEINZER, M. L. (1998) J. Amer. Soc. Brew. Chem., 56,<br />
136.<br />
STEVENS, J. F., TAYLOR, A. W., CLAWSON, J. E. <strong>and</strong> DEINZER, M. L. (1999) J. Agric. Food Chem., 47, 2421.<br />
STEVENS, R. (1967) Chem. Rev., 67, 19.<br />
STEVENS, R. (ed.) (1987) An Introduction to <strong>Brewing</strong> Science <strong>and</strong> Technology. Series II Volume I. Hops.<br />
Institute of <strong>Brewing</strong>, London, pp. vi + 118.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
THORNTON, H. A., KULANDAI, J., BOND, M., JONTEF, M. P., HAWTHORNE, D. B. <strong>and</strong> KAVENAGH, T. E. (1993) J.<br />
Inst. <strong>Brewing</strong>, 99, 473.<br />
TRESSL, R., FRIESE, L., FENDESACK, F. <strong>and</strong> KOPPLER, H. (1978) J. Agric. Food Chem., 26, 1422.<br />
TRESSL, R., ENGEL, K.-H., KOSSA, M. <strong>and</strong> KOPPLER, H. (1983) J. Agric. Food Chem., 31, 892.<br />
VAN ENGEL, E. L. <strong>and</strong> NICKERSON, G. B. (1992) J. Amer. Soc. Brew. Chem., 50, 82.<br />
VERZELE, M. (1958) Bull. Soc. Chim. Belg., 67, 278.<br />
VERZELE, M. <strong>and</strong> DE KEUKELEIRE, D. (1991) Chemistry <strong>and</strong> Analysis of Hop <strong>and</strong> Beer Bitter Acids.<br />
Elsevier, Amsterdam, xx + 417 pp.<br />
WESTWOOD, K. T. <strong>and</strong> DAOUD, I. S. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 579.<br />
WHITEAR, A. L. (1965) Proc. 10th Congr. Eur. Brew. Conv., Stockholm, p. 405.<br />
WHITEAR, A. L. (1966) J. Inst. <strong>Brewing</strong>, 72, 177.<br />
WINDISCH, W., KOLBACH, P. <strong>and</strong> SCHLEICHER R. (1927) Wochschr. Brau., 44, 453.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
9<br />
Chemistry of wort boiling<br />
9.1 Introduction<br />
Wort boiling may be regarded as the turning point in the brewing of beer; it is omitted in<br />
distilling <strong>and</strong> vinegar brewing. At its simplest, in Bavarian <strong>practice</strong>, the all-malt wort is<br />
boiled with hops for 1±2hat atmospheric pressure. Elsewhere, the sweet wort may be<br />
produced from amixed cereal grist <strong>and</strong> additional carbohydrate, either brewing sugars or<br />
wort syrups, may be added to the copper. The boiling vessels were originally made from<br />
copper so are often called coppers although today they are more likely to be made of<br />
stainlesssteel.IntheUSAthewortboilingvesselsarecalledkettles.Aspointedoutinthe<br />
last chapter, the addition of whole hops to the copper is likely to be replaced by hop<br />
pellets or extract. Alternatively, part or all of the hop grist to be added to the copper may<br />
be replaced by post-fermentation isomerized bittering agents <strong>and</strong> hop essences. After the<br />
wort has been boiled with whole hops, the copper is cast into a vessel with a slotted base<br />
called a hop back. A bed of spent hops is formed through which the wort is circulated<br />
until it runs bright, when it is run off, through a heat exchanger, into the fermentation<br />
vessel. Hop pellets <strong>and</strong> extracts do not provide a bed of spent hops for filtration <strong>and</strong> such<br />
worts are usually clarified in awhirlpool (see Chapter 10).<br />
The EBC have published a Manual of Good Practice ± Wort Boiling <strong>and</strong> Clarification<br />
(Denk et al., 2000) in which they list the principal changes that occur during wort boiling<br />
as:<br />
1. Inactivation of malt enzymes<br />
2. Sterilization of the wort<br />
3. Extraction <strong>and</strong> isomerization of compounds derived from hops<br />
4. Coagulation of protein material in the wort<br />
5. Formation of protein/polyphenol complexes<br />
6. Formation of flavour <strong>and</strong> colour complexes<br />
7. Formation of reducing substances to give the wort reducing potential, which is<br />
thought to protect the wort from oxidation later in the process<br />
8. Fall in wort pH<br />
9. Concentration of wort gravity through evaporation of water<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
10. Evaporation of volatile compounds in wort derived from mashing<br />
11. Evaporation of volatile compounds in wort derived from hops.<br />
Other reviews have been provided by Miedaner (1986) <strong>and</strong> O'Rourke (1999, 2002). Of<br />
the changes discussed above, the extraction <strong>and</strong> isomerization of compounds derived<br />
from hops has been discussed in Chapter 8.<br />
Wort boiling requires alot of energy (24±54MJ/hl) <strong>and</strong> the increase in energy prices<br />
over the last few decades has led to many changes in plant design <strong>and</strong> <strong>practice</strong> to<br />
conserve energy (Chapter 10). However, any changes in wort boiling <strong>practice</strong> must be<br />
carefully monitored to ensure they donot affect the quality <strong>and</strong> flavour of the final beer.<br />
The rate of all chemical reactions increases with arise in temperature. Boiling under<br />
pressure increases the temperature <strong>and</strong> so shortens the time necessary to complete the<br />
changes that occur during wort boiling. Low-pressure boiling (c. 1bar of counterpressure)<br />
will raise the temperature to 104±110ëC but not shorten the reaction times<br />
appreciably. However, boiling plants with temperatures of 118±122ëC will require a<br />
holding time of only 8±10min. <strong>and</strong> those with atemperature range of 130±140ëC will<br />
require shorter times with acorresponding saving in energy consumption (see Chapter<br />
10).<br />
9.2 Carbohydrates<br />
During mashing (Chapter 4), malt <strong>and</strong> adjunct polysaccharides, mainly starch, are<br />
enzymatically broken down to simpler units (Table 4.15). Carbohydrates are responsible<br />
for 90% of the extract, 68±75% of which is usually fermentable by yeast. Whole hops<br />
<strong>and</strong> pellets only contribute 0.15% of the total carbohydrates in hopped wort <strong>and</strong> most<br />
hop extracts, prepared with solvents, will contain no carbohydrate. The carbohydrates<br />
undergo little change during wort boiling so the carbohydrate composition of hopped<br />
wort (Table 9.1) is very similar to that of sweet wort. The rate of enzyme-catalysed<br />
reactions also increases with arise in temperature but so does the rate of enzyme<br />
deactivation. Therefore the temperature `optimum' for most enzyme reactions is a<br />
compromise between the two processes. For example, in mashing, the `protein rest' at<br />
50ëC is close to the `optimum' temperature for proteases, peptidases <strong>and</strong> -glucanase<br />
whereas amylases show higher `optima' at 60±70ëC. Above this temperature the rate of<br />
enzyme deactivation dominates so that by 100ëC enzyme activity has ceased <strong>and</strong> the<br />
composition of the wort is fixed. In whisky manufacture, where the wort is not boiled,<br />
enzyme-catalysed reactions continue longer <strong>and</strong> few dextrins survive into the wash.<br />
Similarly, few micro-organisms will survive temperatures of 100ëC. The exceptions are<br />
thermophilic bacteria, mainly Bacillus sp., which form spores that survive wort boiling<br />
(see Chapter 17). However, st<strong>and</strong>ard beer is a poor growth medium for these<br />
organisms..<br />
9.3 Nitrogenous constituents<br />
9.3.1 Introduction<br />
As discussed in Chapter 4, malt contains arange of nitrogenous constituents: proteins,<br />
peptides, <strong>and</strong> amino acids (Section 4.5.1), nucleic acids <strong>and</strong> related substances (Section<br />
4.5.2), miscellaneous substances (Section 4.5.3), <strong>and</strong> vitamins (Section 4.6.1). Many of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 9.1 Carbohydrate composition of worts (results expressed in g/100 ml wort) (MacWilliam, 1968)<br />
a, b<br />
Origin (<strong>and</strong> ref.) Danish<br />
Canadian c<br />
Canadian d<br />
Canadian d<br />
German e<br />
Type of wort Lager Lager Lager Grain lager Lager Pale ale Pale ale<br />
OG 1043.0 1054.0 1048.0 1046.5 1048.5 1040.0 1040.0<br />
Sugar<br />
Fructose 0.21 0.15 0.13 0.10 0.39 0.33 0.97<br />
Glucose 0.91 1.03 0.87 0.50 1.47 1.00<br />
Sucrose 0.23 0.42 0.35 0.10 0.46 0.53 0.60<br />
Maltose 5.24 6.04 5.57 5.50 5.78 3.89 3.91<br />
Maltotriose 1.28 1.77 1.66 1.30 1.46 1.14 1.30<br />
Total ferm. sugar 7.87 9.41 8.58 7.50 9.56 6.89 6.78<br />
Maltotetraose 0.26 0.72 0.54 1.27 0.20 0.53<br />
Higher sugars 2.13* 2.68 2.52 2.94 2.32 1.95<br />
Total dextrins 2.39 3.40 3.06 4.21 2.52 2.48<br />
Total sugars 10.26 12.81 11.64 11.71 9.41 9.26<br />
Sugars (% total extract) 91.1<br />
Fermentability 76.7 73.7 73.7 64.1 73.3 73.2<br />
* The contents of maltopentanose, maltohexaose <strong>and</strong> maltoheptaose in this wort were 0.13, 0.19 <strong>and</strong> 0.18 g/100 ml wort, respectively.<br />
(a) Gjertsen (1953)<br />
(b) Gjertsen (1955)<br />
(c) McFarlane <strong>and</strong> Held (1953)<br />
(d) Latimer et al. (1966)<br />
(e) Kleber et al. (1963)<br />
(f) Harris et al. (1951)<br />
(g) Harris et al. (1954)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
British f<br />
British g
these compounds will be modified during mashing <strong>and</strong> may undergo further changes<br />
during wort boiling. Whole hops <strong>and</strong> pellets will also contain small amounts of these<br />
nitrogenous products.<br />
9.3.2 Proteins<br />
Proteins are macromolecules which may be classified according to their solubility, size<br />
<strong>and</strong> charge (molecular weight <strong>and</strong> isoelectric point), or function, i.e. storage proteins,<br />
structural proteins, <strong>and</strong> proteins that bind other molecules: hormones, immunoproteins,<br />
carrier proteins <strong>and</strong> enzymes. The backbone of all proteins is the polypeptide chain built<br />
up of amino acid residues which is partially broken down by enzymes during mashing.<br />
However, after the enzymes are deactivated, further cleavage of the polypeptide chain<br />
during wort boiling is unlikely as several hours treatment with 6N-hydrochloric acid at<br />
100ëC is necessary for complete hydrolysis of proteins <strong>and</strong> peptides. The primary<br />
structure of aprotein is the sequence of amino acids in the polypeptide chain. These<br />
amino acids (Fig. 4.24) contain other functional groups which can interact to stabilize the<br />
three-dimensional `tertiary' structure of the protein necessary for its biological activity.<br />
Of these cross-linking reactions, the formation of covalent disulphide bridges is very<br />
important. Oxidation of the thiol, -SH, group in cysteine residues gives a disulphide<br />
bridge, -S-S-, as in cystine residues. This reaction is reversible depending on the redox<br />
status of the system. In contrast, the formation of a covalent link between two phenolic<br />
tyrosyl residues is not readily reversible. Aspartic <strong>and</strong> glutamic acid residues in a<br />
polypeptide chain have a free carboxylic acid group (pKa c. 3.5) which can form salts<br />
(ionic bonds) with basic groups found in the side chains of lysine (pKa c. 10), arginine<br />
(pKa 12±13), or histidine (pKa 5.5±7.5).<br />
Hydrogen bonds are also important; hydrogen bonds between the C=O <strong>and</strong> the<br />
-NH- of the peptide bonds can lead to `secondary' structures such as the -helix <strong>and</strong><br />
the -pleated sheet. Hydroxyl groups in serine <strong>and</strong> tyrosine can also form hydrogen<br />
bonds, <strong>and</strong> hydrogen bonds with water will aid the solubility of proteins. In contrast,<br />
the side chains of valine, leucine, isoleucine <strong>and</strong> phenylalanine will produce<br />
hydrophobic areas in the protein structure; such areas may be in the core of the<br />
protein structure. Proteins denature on heating, examples are curdled milk <strong>and</strong> boiled<br />
eggs. As the temperature rises increasing violent thermal motion disrupts the tertiary<br />
structure of the protein, necessary for enzyme activity, <strong>and</strong> hydrophobic groups will<br />
come to the surface of the structure. Here they may interact with other hydrophobic<br />
groups to reduce the solubility of the protein <strong>and</strong> it coagulates. In wort, the protein is<br />
precipitated as the hot break or trub.<br />
The removal of some high molecular weight protein is one of the objects of wort<br />
boiling. Insufficient coagulation <strong>and</strong> the removal of such proteins may effect exchange<br />
processes between yeast cells <strong>and</strong> the surrounding medium (membrane blocking) leading<br />
to an insufficient pH drop in the fermentation. The excess protein may not then be<br />
eliminated during the fermentation <strong>and</strong> lead to clarification problems <strong>and</strong> harsh bitterness<br />
in the final beer. Further, proteins surviving into the final beer may react, on storage, with<br />
polyphenols to form a non-biological haze which will shorten the shelf-life of the beer.<br />
Nevertheless, some proteins are necessary in beer to produce acceptable head retention<br />
<strong>and</strong> mouth-feel.<br />
The coagulation of proteins is strongly influenced by the pH of the wort <strong>and</strong> is most<br />
successful at the isoelectric point of the individual proteins, when the number of positive<br />
<strong>and</strong> negative charges is equal. During wort boiling the pH will drop by 0.1±0.2 pH units<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
to about 5.0. This may be due to the addition of hop bitter acids, the formation of acidic<br />
Maillard products (see later), the precipitation of alkaline phosphates<br />
3Ca 2‡ ‡2HPO 2 4 ˆCa3…PO4† 2 ‡2H ‡<br />
or the reaction of polypeptides with calcium, liberating protons. It is found that above<br />
pH5.0 the amount of nitrogen precipitated during atwo-hour boil is fairly constant but<br />
less is precipitated from more acidic solutions. The amount of hot break formed also<br />
depends on the length <strong>and</strong> vigour of the boil. Even after boiling for three hours more<br />
nitrogen is precipitated if the boil is continued for afurther three hours. Miedaner (1986)<br />
gives the level of coagulable nitrogen in unboiled wort as 35±70ppm which is reduced to<br />
15±25ppm after boiling with a recommended level of 15±18ppm. However, the<br />
measurement of coagulable nitrogen is difficult. It is usually measured as the difference<br />
between the total soluble nitrogen (TSN), determined by Kjeldahl analysis of an aliquot<br />
of the hot water extract, <strong>and</strong> the permanently soluble nitrogen (PSN), remaining after<br />
boiling. It is this last analysis which is not very reproducible.<br />
Worts held at 98±100ëC without boiling or agitation remain turbid which explains the<br />
importance of the vigour of the boil. The intensity of wort circulation depends on copper<br />
design (Chapter 10) <strong>and</strong> evaporation rate. In classical wort boiling at atmospheric<br />
pressure, evaporation of 8±10% of the wort volume/h was recommended, i.e., up to 20%<br />
of the wort volume over two hours. This requires alot of expensive energy <strong>and</strong> to-day<br />
modern kettles operate with a60min. boil with 5±8% evaporation (O'Rourke, 1999).<br />
Reed <strong>and</strong> Jordan (1991) claim that evaporation in excess of 2±3% can be replaced by<br />
suitable agitation. In their laboratory apparatus the fastest clarification was found with<br />
stirringat125rpm.Insufficientorexcessiveagitationincreasesclarificationtime.Itisthe<br />
increase in floc size that is responsible for rapid sedimentation <strong>and</strong> excess agitation will<br />
have ashearing effect on the flocs. The hot break or trub is largely composed of protein.<br />
It was long thought that polyphenols were also involved but protein/polyphenol<br />
complexes, based on hydrogen bonds, are not stable at 100ëC. However, protein/<br />
polyphenol complexes become increasingly stable below 80ëC <strong>and</strong> are found in the cold<br />
break, but this accounts for only 2.5% of the material precipitated (Crompton <strong>and</strong><br />
Hegarty, 1991).<br />
Proteins can be separated according to molecular size by elution from Sephadex gels.<br />
These experiments (Table 9.2) suggest that more high-molecular weight compounds are<br />
precipitated during wort boiling than smaller molecules. Nevertheless, some malt<br />
proteins survive into beer (Sections 4.5.1 <strong>and</strong> 19.1.5). Proteolysis during mashing<br />
produces arange of amino acids (Table 9.3). Whole hops <strong>and</strong> pellets will add small<br />
amounts of these compounds during wort boiling. Small amounts of nitrogen may be lost<br />
as volatile compounds during wort boiling <strong>and</strong> some may be incorporated into<br />
melanoidins but the amino acid spectrum of sweet <strong>and</strong> boiled wort is very similar. The<br />
aminoacidspectrumofbeerisverydifferentasmanyofthesecompoundsaretakenupas<br />
yeast nutrients during fermentation.<br />
Table 9.2 Effect of boiling on the MW distribution of wort proteins (Guenther <strong>and</strong> Stutler, 1965)<br />
Mol. Wt < 5000 5±10,000 10±50,000 50±100,000 > 100,000<br />
Boiled for 95 min. 0.0175 0.0125 0.0040 0.0010 0<br />
Not boiled 0.0336 0.0185 0.0101 0.0023 0.0028<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 9.3 Free amino acids in wort <strong>and</strong> beer (mg/100 ml) (S<strong>and</strong>egren et al., 1954)<br />
Nitrogen <strong>and</strong> amino acids Wort Hopped wort Beer Beer<br />
refermented<br />
Total nitrogen 88.0 84.8 62.6 47.0<br />
Low molecular nitrogen alcohol soluble 63.4 69.5 50.7 35.1<br />
Total -amino nitrogen 42.7 38.0 21.0 13.0<br />
Alcohol soluble -amino nitrogen 37.6 30.8 18.2 2.5<br />
Alanine 9.8 10.2 7.7 1.8<br />
-Amino butyric acid 8.3 7.9 9.6 2.5<br />
Arginine 13.8 5.9 3.0 0.6<br />
Aspartic acid 7.0 9.8 1.6 1.0<br />
Glutamic acid 6.4 3.3 0.8 0.7<br />
Glycine 2.3 2.6 2.1 1.3<br />
Histidine 5.7 3.8 2.8 0.2<br />
Isoleucine 6.2 6.5 2.1 0.3<br />
Leucine 18.1 17.5 4.7 0.7<br />
Lysine 14.9 10.7 2.2 0.5<br />
Phenylalanine 13.7 14.0 4.4 0.6<br />
Proline (imino acid) 45.7 48.3 31.8 33.3<br />
Threonine 5.9 7.3 0.3 0.3<br />
Tyrosine 10.6 9.3 5.9 1.1<br />
Valine 11.9 16.0 6.8 0.4<br />
Serine + Asparagine mM in 100 ml 168.6 171.8 7.9 5.6<br />
Ammonia 2.4 2.4 1.7 1.0<br />
9.4 Carbohydrate-nitrogenous constituent interactions<br />
Sweetwortboiledwithorwithouthopsincreasesincolour.Thisisdueto`non-enzymatic<br />
browning', a reaction between amines, or amino acids, <strong>and</strong> carbonyl compounds,<br />
especially reducing sugars. It is often named after its discoverer Louis-Camille Maillard<br />
(Ikan, 1996, Fayle <strong>and</strong> Gerrard, 2002). The pigments formed by non-enzymatic browning<br />
are called melanoidins <strong>and</strong> the ultimate product is caramel. Non-enzymatic browning<br />
should be distinguished from the action of polyphenoloxidase on substrates such as<br />
tyrosine, which produces the brown or black hair <strong>and</strong> skin pigments called melanins.<br />
In the brewing process the Maillard reaction occurs when malt is kilned <strong>and</strong> continues<br />
during wort boiling. Obviously, dark <strong>and</strong> crystal malts will contain more melanoidins<br />
than pale ale orlagermalts.It was estimatedfor an American beer (presumably pale) that<br />
about one-third of the colour was formed during kilning <strong>and</strong> the other two-thirds during<br />
wort boiling but the ratio will be different using dark malts. In addition to these pigments<br />
the Maillard reaction produces many volatile compounds, some of which have very low<br />
flavour thresholds <strong>and</strong> can influence the flavour of beer. As well as aliphatic compounds,<br />
oxygen, nitrogen <strong>and</strong> sulphur heterocyclic compounds are formed (Mottram, 1994).<br />
There are many reviews of the Maillard reaction (Hodge, 1953; Reynolds, 1963,1965;<br />
Nursten,1980;Waller<strong>and</strong>Feather,1983; Parlimentetal.,1994;Ikan,1996;<strong>and</strong>Fayle<strong>and</strong><br />
Gerrard,2002).ThebasicchemistryisshowninFig.9.1.Theamine,oraminoacid,addsto<br />
the reducing group of the reducing sugar, in the aldehyde form (9.1), to give a product which<br />
is dehydrated to a Schiff's base (9.2). This rearranges to an Amadori compound (an Nsubstituted<br />
1-amino-1-deoxy-2-ketose, (9.3) in the keto form, (9.4) in the enol form).<br />
Amadori compounds characterized in malt include those from: Fru-Ala, Fru-Gly, Fru-Val,<br />
Fru-Leu, Fru-Ile, Fru-Ser, Fru-Thr, Fru-Pyr, Fru-Asp, Fru-Glu, Fru- -aminobutyric acid,<br />
<strong>and</strong> Fru-Pro (Eichner et al., 1994). Fru-Pyr, fructose-pyrrolidonecarboxylic acid, is formed<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Fig. 9.1 The chemistry of non-enzymatic browning.
fromFru-Glu<strong>and</strong>Fru-Glutamineduringworkup.Darkmaltscontainhigherconcentrations<br />
of Amadori products but they cannot be found in malts heated above 200ëC. Considerable<br />
degradation of Amadori products occurs during mashing <strong>and</strong> wort boiling, but apparently<br />
not during fermentation, <strong>and</strong> some products survive into beer. Amadori compounds can<br />
decompose by two routes: 1,2-enolization occurs at low pH values <strong>and</strong> 2,3-enolization at<br />
higherpHvalues.Eichneretal.,(1994)studiedthemodelsystemFru-Glyduringtwohours<br />
inacitratebuffersolution(pH3.0)at90ëC<strong>and</strong>foundthatthemajordecompositionwasvia<br />
3-deoxyglucosone (9.6, R ˆ CH2OH) to 5-hydroxymethylfurfural (HMF, 9.8, R ˆ<br />
CH2OH). The proportion of 3-deoxyglucosone reaches amaximum after 15hafter which<br />
nomoreaccumulates(formationˆdecomposition)buttheconcentrationofHMFcontinues<br />
to increase throughout the reaction. Indeed, the level of HMF can be used to follow the<br />
Maillard reaction.<br />
In beer acorrelation between the stale flavour <strong>and</strong> the HMF content has been reported<br />
(Shimizu et al., 2001). Increased, <strong>and</strong> possibly unacceptable, levels of HMF <strong>and</strong> other<br />
Maillard products are formed during high temperature wort boiling (Fig. 9.2). It is<br />
recommended (Miedaner, 1986) that the temperature should not exceed 140ëC <strong>and</strong> the<br />
holding time should not exceed 2.5min. at 140ëC <strong>and</strong> 3min. at 130ëC. In the same way<br />
pentose sugars will produce furfural (9.8, RˆH, 9.17). Related heterocyclic compounds<br />
found in roasted barley (Harding et al., 1978), malt (Tressl et al., 1977) <strong>and</strong> beer<br />
(Harding et al., 1977; Tressl et al., 1977) are shown in Fig. 9.3±9.7 <strong>and</strong> in Tables 9.4 <strong>and</strong><br />
9.5.Thelevelof2-acetylfuran(9.25),2-acetylthiophene(9.27),furfurylalcohol(9.20),5methylfurfural<br />
(9.22), <strong>and</strong> 5-methylthiophenecarboxaldehyde (9.24)) is higher in ales<br />
than in lagers probably due to the higher kilning temperatures used in the preparation of<br />
themalts. Comparedtopalealemalt,crystal maltcontainsenhancedlevelsofmost ofthe<br />
heterocyclic compounds discussed.<br />
2,3-Enolization of Amadori products gives rise to 1-deoxyosones (9.9). With hexose<br />
sugars, e.g. (9.9, RˆCH2OH), dehydration can give maltol (3-hydroxy-2-methyl-4Hpyran-4-one,<br />
9.11), isomaltol (1-(3-hydroxy-2-furanyl)ethanone, 9.30), <strong>and</strong> 2,5dimethyl-4-hydroxy-3(2H)-furanone<br />
(furaneol, 9.12, RˆCH 3). All these compounds<br />
have sweet caramel flavours <strong>and</strong> maltol (9.11) <strong>and</strong> furfuryl mercaptan (9.29) occur in<br />
beerabove the flavour threshold concentrations.Compounds,such as9.10,which contain<br />
HMF (mg/l)<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0 0 60 120 180 240<br />
Heating time (s)<br />
150°C<br />
140°C<br />
130°C<br />
120°C<br />
Heating time (s)<br />
10 4<br />
10 3<br />
10 2<br />
τ H = f〈σ, [‘HMF’]〉<br />
HMF<br />
25<br />
20<br />
15 mg/l<br />
120 130 140 150<br />
Temperature (°C)<br />
Fig. 9.2 Hydroxymethylyfurfural production during high temperature boiling (Miedaner, 1986).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
X = O<br />
X = NH<br />
X = S<br />
X<br />
CHO<br />
(9.17) Furfural<br />
(9.18) 2-Formylpyrrole<br />
(9.19) 2-Formylthiophene<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
X<br />
(9.20) Furfuryl alcohol<br />
—<br />
(9.21) 2-Hydroxymethylthiophene<br />
CH2OH H3C CHO<br />
X<br />
(9.22) 5-Methylfurfural<br />
(9.23) 2-Formyl-5-methylpyrrole<br />
(9.24) 2-Formyl-5-methylthiophene<br />
O O<br />
O<br />
CO2H CH2SH CO·CH3<br />
(9.28) Furoic acid (9.29) Furfuryl mercaptan<br />
(9.30) Isomaltol<br />
Fig. 9.3 Heterocyclic compounds in roasted barley, wort <strong>and</strong> beer.<br />
CO·CH3<br />
X<br />
(9.25) 2-Acetylfuran<br />
(9.26) 2-Acetylpyrrole<br />
(9.27) 2-Acetylthiophene<br />
OH
R O<br />
C<br />
R'<br />
R<br />
C<br />
C<br />
O<br />
C<br />
R' O<br />
NH2<br />
+<br />
H2N<br />
R N R'<br />
R' N R<br />
R<br />
R'<br />
+ R"·CH<br />
NH2<br />
COOH<br />
α-Diketone α-Amino acid<br />
α-Aminoketone<br />
N<br />
N<br />
[ – O]<br />
R'<br />
R<br />
Alkylpyrazine<br />
O<br />
+ R"·CHO<br />
Strecker<br />
aldehyde<br />
C<br />
C<br />
R'<br />
R<br />
+ H2O<br />
the -C(OH) ˆC(OH).CˆO grouping are called reductones <strong>and</strong> combine with oxygen to<br />
maintain the oxidation/reduction (redox) balance of asystem. Another example of a<br />
reductone is ascorbic acid (vitamin C). Fragmentation of the deoxyosone intermediates<br />
can lead to the -dicarbonyl compounds, pyruvaldehyde (9.13) <strong>and</strong> 2,3-butanedione<br />
(diacetyl, 9.14) <strong>and</strong> the related ketols hydroxyacetone (9.15) 3-hydroxy-2-butanone<br />
(acetoin, 9.16).<br />
The deoxyosones <strong>and</strong> other -dicarbonyl compounds can react with amino acids<br />
according to the Strecker reaction (Fig. 9.4) to give aldehydes, with one carbon less than<br />
the amino acids, carbon dioxide, <strong>and</strong> an -amino ketone. Most of the Strecker aldehydes<br />
(Table 9.6) have potent aromas but those that survive wort boiling are likely to be<br />
reduced to the corresponding alcohols during fermentation. The amino ketones formed in<br />
the Strecker reaction can condense together to form, after oxidation, pyrazines (9.34).<br />
The pyrazines found in roasted barley malt <strong>and</strong> beer are listed in Table 9.6. As illustrated<br />
in Fig. 9.4, the amino acid cysteine breaks down in the Strecker reaction to liberate<br />
O<br />
R"<br />
C<br />
O<br />
H<br />
H<br />
R"<br />
N<br />
O<br />
N<br />
O<br />
– CO2<br />
C<br />
C<br />
C<br />
R<br />
C R'<br />
R<br />
R'<br />
For cysteine (R" = HS·CH )<br />
CH3CHO + NH3 + H2S<br />
Fig. 9.4 Strecker reaction of -amino acids <strong>and</strong> formation of alkylpyrazines.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
R<br />
R'<br />
+<br />
C<br />
C<br />
O<br />
O<br />
2–
N<br />
H<br />
H<br />
COOH<br />
(4.44) Proline α-Diketone<br />
N<br />
H<br />
N N<br />
CO·CH3<br />
(9.36) Pyrrolidine (9.37) 1-Pyrroline (9.38) 2-Acetyl-1-pyrroline<br />
N<br />
H<br />
N<br />
CO·CH3<br />
(9.40) 2-Acetyl-1,4,5,6tetrahydropyridine<br />
CO·CH3<br />
+<br />
hydrogen sulphide, which is probably incorporated into the sulphur heterocyclic<br />
compounds formed during wort boiling. The imino acid proline is not strictly an amino<br />
acid but is the major `amino acid' in wort (Table 9.3). In the Strecker reaction it cannot<br />
form a Strecker aldehyde <strong>and</strong> an -amino ketone but it reacts with -dicarbonyl<br />
compounds to form nitrogen heterocyclic compounds such as 1-pyrroline (9.37),<br />
pyrrolidine (9.36), 1-acetonyl-2-pyrroline (9.39), <strong>and</strong> 2-acetyl-1,4,5,6-tetrahydropyridine<br />
(9.40) (Fig. 9.5).<br />
Roberts <strong>and</strong> Acree (1994) studied the Glu-Pro Malliard reaction by GC-Olfactometry.<br />
The seven most potent products, characterized by olfactometry, were: diacetyl (9.14,<br />
%`Charm', 0.5), 2-acetyl-1-pyrroline(9.38, 19), 2-acetyl-1,4,5,6-tetrahydropyridine<br />
(9.40, 12), 2-acetylpyridine (9.42, 19), 2-acetyl-3,4,5,6-tetrahydropyridine (9.41, 63),<br />
furaneol (9.12, RˆCH3, 3.9) <strong>and</strong> 5-acetyl-2,3-1H-pyrrolizine (9.43, 0.3). According to<br />
O<br />
O<br />
C<br />
C<br />
N<br />
O<br />
(9.39) 1-Acetonyl-2-pyrroline<br />
CO·CH3 CO·CH3<br />
N N<br />
(9.41) 2-Acetyl-3,4,5,6tetrahydropyridine<br />
O<br />
(9.42) 2-Acetylpyridine<br />
(9.43) 5-Acetyl-2,3-1H-pyrrolizine (9.44) Maltoxacine<br />
Fig. 9.5 Products of the Strecker reaction with proline.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
N<br />
O
SH<br />
NH2<br />
(9.45)<br />
Cysteamine<br />
+ H<br />
O<br />
O<br />
(9.13)<br />
Pyruvaldehyde<br />
Flavour profile score<br />
1.0<br />
0.8<br />
0.6<br />
0.4<br />
0.2<br />
N<br />
H<br />
S<br />
CO·CH3<br />
(9.46)<br />
2-Acetylthiazolidine<br />
N<br />
the mass yield this last compound was the major product of the reaction. Maltoxazine<br />
(9.44) was also formed in considerable amounts but had little or no aroma. Details of<br />
Maillard reactions between most of the amino acids (Table 9.3) <strong>and</strong> the common sugars<br />
can be found in the literature.<br />
The presence of thiazoles in roasted barley, malt <strong>and</strong> beer is recorded in Table 9.4. Of<br />
particular interest is 2-acetyl-2-thiazoline (9.47) <strong>and</strong> 2-acetylthiazole (9.48) which have<br />
popcorn flavours with low thresholds. The former compound is thought to be formed<br />
(Fig. 9.6) by condensation of cysteamine (9.45) (formed by decarboxylation of cysteine)<br />
<strong>and</strong> 2-oxopropanal (9.13, pyruvaldehyde) to give 2-acetylthiazolidine (9.46) which, by<br />
dehydrogenation, gives (9.47) <strong>and</strong> probably (9.48). A Maillard reaction between<br />
cysteamine <strong>and</strong> fructose produces N-(2-mercaptoethyl)-1,3-thiazolidine (9.49) which has<br />
an intense popcorn-like odour with an extremely low threshold (0.005ng/l in air) (Engel<br />
<strong>and</strong>Schieberle,2002).Thiazolesmayalsobeformedbydegradationofthiamine(vitamin<br />
B1, 4.68).<br />
As well as these Maillard reaction products, malt <strong>and</strong> hops will contribute volatile<br />
compounds to wort which, if not partially removed, will lead to unacceptable beers.<br />
Normally the excess of these compounds is lost by evaporation during wort boiling. As<br />
– 2H<br />
N<br />
S<br />
CO·CH3<br />
(9.47)<br />
2-Acetyl-2-thiazoline<br />
SH<br />
(9.49) N-(2-Mercaptoethyl)-1,3-thiazolidine<br />
S<br />
Fig. 9.6 Formation of thiazoles.<br />
Sulphidic<br />
Grassy/grainy<br />
0 0 1 2 3 4 5<br />
Evaporation (%)<br />
Fruity<br />
– 2H<br />
N<br />
S<br />
(9.48)<br />
2-Acetylthiazole<br />
Fig. 9.7 Effect of evaporation on some important flavour characters (O'Rourke, 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
CO·CH3
Table 9.4 Derivatives of pyrrole, thiazole <strong>and</strong> pyridine in roasted barley, malt <strong>and</strong> beer (see text<br />
for references)<br />
(9.31) Pyrrole (9.32) Thiazole (9.33) Pyridine<br />
Unsubstituted Unsubstituted Unsubstituted<br />
2-Methyl 4-Methyl Methyl<br />
5-Methyl<br />
Dimethyl<br />
1-Acetyl<br />
2-Acetyl (9.26) 2-Acetyl 2-Acetyl<br />
3-Acetyl<br />
2-Formyl-1-methyl<br />
2-Formyl-5-methyl<br />
1-Ethyl-2-formyl<br />
1-Furfuryl<br />
5-Hydroxyethyl-4methyl<br />
Table 9.5 Pyrazines in roasted barley, malt <strong>and</strong> beer (for references see text)<br />
(9.34) Pyrazine (9.35) Cyclopentapyrazine<br />
Pyrazine Concentration (ppb)<br />
Malt Beer<br />
1. Methylpyrazine 280 70<br />
2. 2, 5-Dimethylpyrazine 130 110<br />
3. 2, 6-Dimethylpyrazine 120 35<br />
4. 2, 3-Dimethylpyrazine 200 15<br />
5. Ethylpyrazine 140 10<br />
6. 2-Ethyl-6-methylpyrazine 80 }35<br />
7. 2-Ethyl-5-methylpyrazine 40 }<br />
8. 2-Ethyl-3-methylpyrazine 80 +<br />
9. Trimethylpyrazine 320 20<br />
10. 2-Ethyl-3, 6-dimethylpyrazine 10 20<br />
11. 2-Ethyl-3, 5-dimethylpyrazine 30 10<br />
12. 2-Ethyl-5, 6-dimethylpyrazine 10 +<br />
13. Tetramethylpyrazine 110 +<br />
14. 6, 7-Dihydro-5H-cyclopentapyrazine + 10<br />
15. 5-Methyl-6, 7-dihydro-5H-cyclopentapyrazine 20 15<br />
16. 2-Methyl-6, 7-dihydro-5H-cyclopentapyrazine 20 10<br />
17. 5-Methylcyclopentapyrazine 10 +<br />
18. 2-Furfurylpyrazine + 25<br />
19. 2-(2'-Furfuryl)methylpyrazine 10<br />
20. 2-(2'-Furfuryl)dimethylpyrazine +<br />
+ Detected but not quantified<br />
Not detected<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 9.6 Aldehydes produced in the Strecker reaction (after Ho, 1996)<br />
Amino acid Aldehyde Odour properties Flavour<br />
threshold<br />
(ppm) a<br />
Alanine Acetaldehyde Pungent, ethereal, green, sweet 10<br />
CH3.CHO<br />
Valine Isobutyraldehyde Extremely diffusive, pungent, green, (1.0)<br />
(CH3)2CH.CHO in extreme dilution almost pleasant,<br />
fruity, banana-like<br />
Leucine Isovaleraldehyde Very powerful acrid-pungent. In (0.6)<br />
(CH 3) 2CH.CH 2CHO extreme dilution fruity, rather pleasant<br />
Isoleucine 2-Methylbutanal Powerful, but in extreme dilution 1.25<br />
CH3CH2CH(CH3)CHO almost fruity ± `fermented' with a<br />
peculiar note resembling that of<br />
roasted cocoa or coffee<br />
Methionine Methional Powerful onion-meat-like, potato-like (0.25)<br />
CH3S.CH2CH2CHO<br />
Phenylalanine Phenylacetaldehyde Very powerful, pungent, floral <strong>and</strong> sweet (1.6)<br />
C6H5CH2CHO<br />
a Meilgaard, 1975<br />
shown in Fig. 9.7 grassy/grainy <strong>and</strong> sulphidic aromas are greatly reduced with only 2%<br />
evaporation. An important malt-derived volatile is dimethyl sulphide (DMS, 4.112), the<br />
flavour threshold of which is 40±60ppb but some all-malt lagers with 100ppb DMS are<br />
found acceptable. Above this level DMS gives asweetcorn flavour. DMS is produced by<br />
thermal decomposition of S-methylmethionine (Fig. 4.34), the half-life of which is<br />
reported to be 35min. at 100ëC. DMS formed by kilning <strong>and</strong> wort boiling will be rapidly<br />
lost by evaporation but S-methylmethionine will continue to break down during wort<br />
cooling <strong>and</strong> the DMS formed then will persist into beer. To minimize such DMS<br />
formation it is recommended (O'Rourke, 1999, 2002) to use malts with low Smethylmethionine<br />
contents <strong>and</strong> to extend the wort boiling time to decompose the<br />
majority of the precursor <strong>and</strong> drive off the DMS. Worts from high-temperature wort<br />
boiling systems contain negligible amounts of DMS <strong>and</strong> its precursors. It is also<br />
recommended to minimize the whirlpool st<strong>and</strong> time <strong>and</strong> to use quick wort cooling to<br />
reduce the time that the wort is held hot.<br />
When wort is boiled with whole hops or pellets, the majority of the hop oil<br />
constituents will be lost during a60±90min. boil in an open copper. If late hop character<br />
is required aportion (up to 20%) of the hop gristmay be added, as choice aroma hops, 5±<br />
15min. before the end of the boil. Early attempts at high temperature wort boiling, with<br />
insufficient venting, produced worts with unacceptable levels of hop oils. Excess<br />
Maillard volatile products must also be evaporated. Figure 9.8 shows the amounts of<br />
various heterocyclic compounds in the vapour condensate during wort boiling. Of<br />
particular interest is 2-acetylthiazole, which has a flavour threshold of 10 ppb in beer, <strong>and</strong><br />
must be reduced if not to cause an off-flavour.<br />
9.4.1 Melanoidins<br />
The volatile products of the Maillard reaction have been studied in more detail than the<br />
melanoidin pigments. These are obviously heterogeneous depending on the sugars <strong>and</strong><br />
amino acids involved, their ratios <strong>and</strong> the pH <strong>and</strong> temperature of the reaction (Ikan, 1996).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
μg per 6l condensate<br />
300<br />
200<br />
100<br />
2-Ac-thiazole<br />
2-Ac-pyridine<br />
2-Ac-pyrrole<br />
2-Mc-pyrazine<br />
0 0 20 40 60 80 100 120<br />
Boiling time (min.)<br />
Fig. 9.8 N-Heterocyclic compounds in the evaporation condensate during wort boiling<br />
(Miedaner, 1986).<br />
At neutral or slightly acidic pH, osones, furfurals, <strong>and</strong> probably other heterocyclic<br />
compoundsare involved inthe formation of melanoidins, which are highmolecular weight<br />
compounds (10,000±30,000 daltons by ultracentrifugation; 20,000 by gel filtration).<br />
Isoelectric focusing electrophoresis of the melanoidins from aGlu-Gly reaction showed 20<br />
b<strong>and</strong>s while those from aXyl-Gly reaction showed 16 b<strong>and</strong>s. Spectroscopic studies of the<br />
products, with <strong>and</strong> without isotopic labelling, <strong>and</strong> ozonolysis has led to the suggestion that<br />
the melanoidins contain the repeating structure shown in Fig. 9.9. The melanoidins, which<br />
may resemble the humic acids in soil, show many interesting biological reactions which<br />
may influence the brewing process (Ikan, 1996). For example, they show antimicrobial<br />
activity, especially towards enteric bacteria; they inhibit trypsin <strong>and</strong> have dietary fibre-like<br />
action; they show mutagenic action; they react with metal ions <strong>and</strong> show antioxidative<br />
effects. In particular, melanoidins scavenge hydroxyl radicals, hydrogen peroxide <strong>and</strong><br />
superoxides (Hayase, 1996) <strong>and</strong> have an influence on beer foam.<br />
9.4.2 Caramel<br />
TheFAO/WHO<strong>and</strong>theEECScientific Committeeforfoodhaveacceptedfourclassesof<br />
caramel (Thornton, 1989):<br />
ClassI. Caramelpreparedbythecontrolledheattreatmentofcarbohydrateswithor<br />
without the presence of food quality alkali or acid.<br />
Class II. Caramel prepared by the controlled heat treatment of carbohydrates with<br />
caustic sulphites.<br />
Class III. Caramel prepared by the controlled heat treatment of carbohydrates with<br />
ammonia.<br />
Class IV. Caramel prepared by the controlled heat treatment of carbohydrates with<br />
ammonium <strong>and</strong> sulphite containing compounds.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH O CH O<br />
CH R<br />
C N R C N R<br />
HC N CH R<br />
CH2<br />
HC OH<br />
CH<br />
HC OH<br />
O<br />
CH C<br />
C OH CH<br />
N CH2<br />
C<br />
R<br />
N CH2 R<br />
HC<br />
R'<br />
OH HC<br />
R'<br />
OH<br />
CH<br />
R'<br />
C<br />
HC<br />
OH<br />
OH<br />
O<br />
CH2<br />
C O<br />
C<br />
CH<br />
N<br />
R' CH C OH<br />
R – NH2 = amine<br />
R' C OH<br />
R' = H or CH2OH R'<br />
^<br />
n<br />
Fig. 9.9 Possible repeating units of melanoidins <strong>and</strong> their precursors<br />
(after Kato <strong>and</strong> Tsuchida, 1981).<br />
Class Icaramels are used inspirits<strong>and</strong>liqueurs, Class IIcaramels invermouths <strong>and</strong>other<br />
aperitifs, <strong>and</strong> Class IV caramels in soft drinks. Class III caramels, ammonia caramels, are<br />
used in bakery <strong>and</strong> meat products <strong>and</strong> in brewing. Ammonia caramel (E 150 (c)) is used<br />
for colouring/colour adjustmentofbeer <strong>and</strong>istheonlycolouring matter permittedinbeer<br />
in the UK. These electropositive caramels are prepared from glucose syrups, with high<br />
dextroseequivalents,<strong>and</strong>0.880ammonia.Theyareallowedtoreacttogetherforatleasta<br />
week at ambient temperature, then at 90ëC overnight, <strong>and</strong> finally at 120ëC for about<br />
three hours. Heating must be carefully controlled to maintain abalance between colour<br />
<strong>and</strong> viscosity. At the appropiate time the mixture is cooled to 80ëC, softened water added<br />
<strong>and</strong> the product blended as required.<br />
The product contains about 25% extract <strong>and</strong> 32,000±48,000 EBC colour units (see<br />
Chapter 19). The UK Food Advisory Committee has recommended that there should be a<br />
limit for caramel in beer of 5000 mg/kg. Other committees have established an acceptable<br />
daily intake of ammonia caramel of 200 mg/kg bodyweight for man. The EEC<br />
specification states that not more than 50% of the colour should be bound by DEAE<br />
cellulose but more than 50% should be bound by phosphoryl cellulose. Caramel contains<br />
0.7±3.3% total nitrogen but not more than 0.3% ammoniacal nitrogen <strong>and</strong> not more than<br />
0.2% total sulphur. The EEC also proposes limits for two possible by-products of caramel<br />
production (Fig. 9.10): not more than 250 mg/kg of 4-methylimidazole (9.50) <strong>and</strong> not<br />
more than 10 mg/kg of 2-acetyl-4-tetrahydroxybutylimidazole (9.51). There are the usual<br />
limits for heavy metals, etc. Most commercial caramels will comply with these limits.<br />
Class IV electronegative caramels cannot be used in beer as they react with the<br />
electropositive finings. Caramel can be added to the copper but is usually added with<br />
primings to make minor adjustments to the colour of beer. However, the fundamental<br />
colour of beer will be determined by the choice of malt <strong>and</strong> adjuncts added to the copper.<br />
H3C<br />
N<br />
H<br />
N HO·CH2·CH(OH)·CH(OH)·CH(OH) N<br />
N<br />
H<br />
CO·CH3<br />
(9.50) 4-Methylimidazole (9.51) 2-Acetyl-4-tetrahydroxybutylimidazole<br />
Fig. 9.10 Imidazoles limited in a caramel.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
9.5 Protein-polyphenol (tannin) interactions<br />
The polyphenols in malt (Section 4.4.9) <strong>and</strong> hops (Section 8.4) have been discussed earlier<br />
(see pp. 158±60; 302 for structures). The phenolic acids (hydroxybenzoic acids <strong>and</strong><br />
hydroxycinnamic acids) <strong>and</strong> the hop flavonols (usually as glycosides) readily dissolve<br />
during mashing <strong>and</strong> wort boiling <strong>and</strong> most survive into beer. McMurrough <strong>and</strong> Delcour<br />
(1994) regard the flavanoids (proanthocyanidins) as the most important polyphenols in<br />
wort <strong>and</strong> classify them as: simple flavanols ((+)-catechin, ( )-epicatechin, dimeric <strong>and</strong><br />
trimeric flavanoids); polymeric flavanols (`oxidation' products of the simple flavanoids,<br />
<strong>and</strong> complex flavanols (in water-soluble association with polypeptides).<br />
Polyphenols form strong hydrogen bonds with polypeptides <strong>and</strong> these complexes are<br />
comparatively inert. As already mentioned they dissociate above 80ëC <strong>and</strong> the<br />
polypeptide can be displaced with urea, H2N.CO.NH2. They are non-tanning <strong>and</strong> their<br />
presence in beers does not necessarily lead to haze formation, even in beers stored in<br />
bottles with excessive headspace air. Simple flavanoids are very susceptible to free<br />
radical reactions initiated either by oxygen or plant oxidase/peroxidase enzymes. These<br />
reactions can lead to cross linking, an increase in molecular size, the formation of aredbrown<br />
colouration <strong>and</strong> the formation of polymeric flavanols which are strongly tanning.<br />
The tanning properties of the simple flavanoids increase with increases in the molecular<br />
weight.<br />
Wort boiling causes abig change in its phenolic make-up. In sweet wort the dimeric<br />
flavanols (prodelphinidin B3 <strong>and</strong> procyanidin B3, 18mg/l) <strong>and</strong> (+)-catechin (6mg/l)<br />
predominate.Afterboilingwithhopsthelevelofdimeric<strong>and</strong>trimericflavanolsfalls(to6<br />
<strong>and</strong> 5mg/l respectively) while that of (+)-catechin <strong>and</strong> ( )-epicatechin increases (to<br />
10mg/l). The ( )-epicatechin is derived not only from hops but also by epimerization of<br />
(+)-catechin. Derdelinckx <strong>and</strong> Jerumanis (1987) also showed that proanthocyanidins<br />
depolymerize during brewing. McMurrough <strong>and</strong> Delacour (1994) conclude that it is the<br />
simple flavanols that are the haze precursors in wort.<br />
Similarly, Whittle et al. (1999) studied the polyphenols in barley <strong>and</strong> beer. They<br />
identified over 50 flavanols in barley, including seven pentamers, but in beer only 24<br />
flavanols were found including (+)-catechin <strong>and</strong> ( )-epicatechin. No tetramers <strong>and</strong><br />
pentamers survived into beer but it was not established whether they were destroyed in<br />
mashing or in wort boiling. However, these flavanols do not appear to be necessary for<br />
protein coagulation. Beers were made with aproanthocyanidin-free malt (Galant) <strong>and</strong> a<br />
regularmalt (Triumph)<strong>and</strong> bittered either with Saaz hops oratannin-freehexane extract.<br />
All four brews showed similar levels of coagulatable nitrogen after a60 minute boil.<br />
After alonger boil (90 min.) the coagulatable nitrogen was reduced further in every case<br />
(Delacour et al., 1988). Techniques to remove protein-polyphenol complexes from beer<br />
are discussed in Chapter 15.<br />
9.6 Copper finings <strong>and</strong> trub formation<br />
The importance of removing some of the protein from sweet wort <strong>and</strong> the difficulties that<br />
can arise later if this is not accomplished have been mentioned <strong>and</strong> is discussed further in<br />
Chapter 15. With a view to improving trub formation many brewers add electronegative<br />
finings to the copper at or near the end of the boil (4±8 g/hl). These copper finings are<br />
usually Irish moss, the dried red marine algae Chondrus cripus (plus some Gigartina<br />
stellata) or the purified polysaccharide therefrom, -(kappa)-carrageenan. This<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Addition rate<br />
Sediment<br />
Haze<br />
Fig. 9.11 The effect of copper fining rate on performance (Leather, 1998).<br />
polysaccharide is made up of achain of galactose <strong>and</strong> occasionally anhydrogalactose,<br />
units linked alternately 1±3 <strong>and</strong> 1±4. Some of the free hydroxyl groups are esterified with<br />
sulphate groups providing the negative charge. Instead of Irish moss, some brewers add<br />
silica gel. These copper finings should be distinguished from the proteinaceous isinglass<br />
finings usedtoclarify beereither incask orconditioningtank(Chapter 15).Nevertheless,<br />
the use of copper finings reduces the non-microbiological particles (NMP) in the final<br />
beer (Leather et al., 1996).<br />
Under optimum boiling conditions, e.g., avigorous rolling boil under atmospheric<br />
pressure at 102ëC for at least an hour, the hot break is formed as large flocs which can be<br />
removedinthehopbackorwhirlpool.Whenthewortisboiledwithwholehopsorpellets<br />
the trub adheres to the hop debris. In contrast, with inefficient wort boiling, or in aplant<br />
producing excessive shear, the trub may separate as fine flocs, which remain in<br />
suspension. Similarly, the cold break consists of very fine particles that are slow to form<br />
<strong>and</strong> settle <strong>and</strong> consequently may survive fermentation <strong>and</strong> be carried over into the beer<br />
(Chapter 10). Leather (1998) lists eleven factors which affect copper fining performance.<br />
As the dose rate increases, wort clarity improves <strong>and</strong> the amount of sediment also<br />
increases (Fig. 9.11). The optimum fining rate is that which produces the best wort clarity<br />
with the minimum volume of sediment. In <strong>practice</strong> this produces a beer containing<br />
approximately 10 6 non-microbiological particles/ml in each of the three size fractions,<br />
< 2 m, 2±10 m, <strong>and</strong> > 10 m. The optimum fining rate is found by experiment. Wort<br />
clarity is assessed on an arbitary scale from A (brilliant) to F (cloudy) or can be measured<br />
by the absorbance at 600 nm against a membrane filtered wort.<br />
Copper finings have no significant effect on hot wort clarity, their main effect being<br />
the production of bright cold wort. They are added to hot wort since -carrageenan does<br />
Table 9.7 Typical composition of hot <strong>and</strong> cold break (Moll <strong>and</strong> de Blauwe, 1994)<br />
Hot break Cold break<br />
Particle size mm 30±80 0.5±1.0<br />
Typical wet weight g/hl 150±400 5±30<br />
Moisture % 73±85 70±80<br />
Proteinaceous matter degree 40±70 45±75<br />
Bittering substances % 10±20 n/a<br />
Polyphenols % 5±10 10±30<br />
Carbohydrates % 4±8 20±30<br />
Ash % 3±5 2±3<br />
Fats % 1±2 n/a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
not dissolve below 60ëC. The copper finings should be added early enough in the boil so<br />
that they all dissolve but late enough so that they are not significantly degraded. For the<br />
same reason long st<strong>and</strong>s in the whirlpool should be avoided. Awort of pH of c. 5.0 is<br />
required for efficient fining <strong>and</strong> worts below pH 4.5 often fail to fine. Adifference of 0.3<br />
pH units can make adifference between optimum (A) clarity <strong>and</strong> poor (D) clarity. Malt<br />
variety <strong>and</strong> quality also influences copper fining performance. As well as the influence<br />
malt has on the pH of the wort, the amount of cold break protein (the amount of cold<br />
break that forms naturally without the addition of copper finings) correlates positively<br />
with copper fining performance.<br />
Analyses of hot <strong>and</strong> cold break are given in Table 9.7. It will be noted that in terms of<br />
mass the cold break is less than 20% of the hot break <strong>and</strong> that both are largely made up of<br />
proteinaceous materials. The cold break is richer in polyphenols <strong>and</strong> carbohydrates.<br />
Although hydrogen bonds between polyphenols <strong>and</strong> polypeptides are largely dissociated<br />
above 80 ëC, some malt <strong>and</strong> hop polyphenols are precipitated with the hot break but,<br />
according to McMurrough <strong>and</strong> Delacour (1994), `their role is more passive than active'.<br />
Ninety per cent of the lipids in the copper are deposited with the trub <strong>and</strong> spent hops<br />
(Anness <strong>and</strong> Reed, 1985). In addition some fatty acids may be lost by steam<br />
volatilization; with 12% evaporation 0.9% of the lipids in wort were lost by this route.<br />
9.7 References<br />
ANNESS, B. J. <strong>and</strong> REED, R. J. R. (1985) J. Inst. <strong>Brewing</strong>, 91, 82.<br />
CROMPTON, I. E. <strong>and</strong> HEGARTY, P. K. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 625.<br />
DELCOUR, J. A., VANHAMEL, S., MOERMAN, E. <strong>and</strong> VANCRAENENBROECK, R. (1988) J. Inst. <strong>Brewing</strong>, 96,<br />
371.<br />
DENK, V., FELGENTRAEGER, H. G. W., FLAD, W., LENEOL, M., MICHEL, R., MIEDANER, H., STIPPLER, K.,<br />
HENSEL, H., NARZISS, L. <strong>and</strong> O'ROURKE, T. (2000) European Brewery Convention ± Manual of Good<br />
Practice, Wort Boiling <strong>and</strong> Clarification, pp. xvi + 176. Fachverlag Hans Carl, NuÈrnberg.<br />
DERDELINCKX, G. <strong>and</strong> JERUMANIS, J. (1987) Proc. 21st Congr. Eur. Brew. Conv., Madrid, p. 577.<br />
EICHNER, K., REUTTER, M. <strong>and</strong> WITTMANN, R. (1994) in Parliment, T. H., Morello, M. J. <strong>and</strong> McGorrin, R.<br />
J. (eds) Thermally Generated Flavors ± Maillard, Microwave <strong>and</strong> Extrusion Processes. ACS<br />
Symposium Series No. 543, American Chemical Society, Washington DC, p. 42.<br />
ENGEL, W. <strong>and</strong> SCHIEBERLE, P. (2002) J. Agric. Food Chem., 50, 5391.<br />
EUROPEAN BREWERY CONVENTION See Denk et al.<br />
FAYLE, S. E. <strong>and</strong> GERRARD, J. A. (2002) The Maillard Reaction, Royal Society of Chemistry, London, xiv<br />
+ 120 pp.<br />
GJERTSEN, P. (1953) J. Inst. <strong>Brewing</strong>, 59, 296.<br />
GJERTSEN, P. (1955) Proc. 5th Congr. Eur. Brew. Conv., Baden-Baden, p. 37.<br />
GUENTHER, K. R. <strong>and</strong> STUTLER, J. R. (1965) Proc. Annu. Meet. Amer. Soc. Brew. Chem., p. 30.<br />
HARDING, R. J., NURSTEN, H. E. <strong>and</strong> WREN, J. J. (1977) J. Sci. Food Agric., 28, 225.<br />
HARDING, R. J., WREN, J. J. <strong>and</strong> NURSTEN, H. E. (1978) J. Inst. <strong>Brewing</strong>, 84, 31.<br />
HARRIS, G., BARTON-WRIGHT, E. C. <strong>and</strong> CURTIS, N. (1951) J. Inst. <strong>Brewing</strong>, 57, 264.<br />
HARRIS, G., HALL, R. D. <strong>and</strong> MACWILLIAM, I. C. (1954) J. Inst. <strong>Brewing</strong>, 60, 464.<br />
HAYASE, F. (1996) in Ikan, R. (ed.) The Maillard Reaction. John Wiley, Chichester, p. 89.<br />
HO, C.-T. (1996) in Ikan, R. (ed.) The Maillard Reaction. John Wiley, Chichester, p. 27.<br />
HODGE, J. E. (1953) J. Agric. Food Chem., 1, 928.<br />
IKAN, R. (1996) The Maillard Reaction. John Wiley, Chichester, p. 228.<br />
KATO, H. <strong>and</strong> TSUCHIDA, H. (1981) Prog. Food Nutr. Sci., 5, 147.<br />
KLEBER, W., SCHMID, P. <strong>and</strong> SEYFARTH, I. (1963) Brauwissenschaft, 16, 1.<br />
LATIMER, R. A., LAKSHMINARAYANAN, K., QUITTENTON, R. C. <strong>and</strong> DENNIS, G. E. (1966) Proc Conv. Inst.<br />
Brew. Australian Section, p. 111.<br />
LEATHER, R. V. (1998) J. Inst. <strong>Brewing</strong>, 104, 9.<br />
LEATHER, R. V., WARD, I. L., MORSON, B. T. <strong>and</strong> DALE, C. J. (1996), Ferment, 9, 31.<br />
MCFARLANE, W. D. <strong>and</strong> HELD, H. R. (1953) Proc. 4th Congr. Eur. Brew. Conv., Nice, p. 110.<br />
MCMURROUGH, I. <strong>and</strong> DELCOUR, J. A. (1994) Ferment, 7, 175.<br />
MACWILLIAM, I. C. (1968) J. Inst. <strong>Brewing</strong>, 74, 38.<br />
MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 151.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
MIEDANER, H. (1986) J. Inst. <strong>Brewing</strong>, 92, 330.<br />
MOLL, M. <strong>and</strong> DE BLAUWE, J. J. (1994) Beers <strong>and</strong> Coolers, (trans. by Wainwright, T.), Intercept, Andover.<br />
495 pp.<br />
MOTTRAM, D. S. (1994) in Parliment, T. H., Morello, M. J. <strong>and</strong> McGorrin, R. J. Thermally Generated<br />
Flavors ± Maillard, Microwave <strong>and</strong> Extrusion Processes. ACS Symposium Series No. 543.<br />
American Chemical Society, Washington DC, p. 104.<br />
NURSTEN, H. E. (1980) Food Chem., 6, 263.<br />
O'ROURKE, T. (1999) Brewers Guardian, 128 (8), 34; 128 (9), 38.<br />
O'ROURKE, T. (2002) The Brewer International, 2 (2), 17.<br />
PARLIMENT, T. H., MORELLO, M. J. <strong>and</strong> MCGORRIN, R. J. (eds) (1994) Thermally Generated Flavors ±<br />
Maillard, Microwave <strong>and</strong> Extrusion Processes. ACS Symposium Series No. 543. American<br />
Chemical Society, Washington DC, pp. x + 492.<br />
REED, R. J. R. <strong>and</strong> JORDAN, G. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 673.<br />
REYNOLDS, T. M. (1963) Advances in Food Research, 12, 1.<br />
REYNOLDS, T. M. (1965) Advances in Food Research, 14, 167.<br />
ROBERTS, D. D. <strong>and</strong> ACREE, T. E. (1994) in Parliment, T. H., Morello, M. J. <strong>and</strong> McGorrin, R. J. Thermally<br />
Generated Flavors ± Maillard, Microwave <strong>and</strong> Extrusion Processes. ACS Symposium Series No.<br />
543. American Chemical Society, Washington DC, p. 71.<br />
SANDEGREN, E., ENEBO, L., GUTHENBERG, H. <strong>and</strong> LJUNGDAHL, L. (1954) Proc. Annu. Meet. Amer. Soc.<br />
Brew. Chem., p. 63.<br />
SHIMIZU, C., NAKAMURA, Y., MIYAL, K., ARAKI, K., TAKASHIO, W. <strong>and</strong> SHINOTSUKA, K. (2001) J. Amer. Soc.<br />
Brew. Chem., 49, 51.<br />
THORNTON, J (1989) <strong>Brewing</strong> <strong>and</strong> Distilling International, 20 (10), 36.<br />
TRESSL, R., RENNER, R., KOSSA, T. <strong>and</strong> KOPPLER, H. (1977) Proc. 16th Congr. Eur. Brew. Conv.,<br />
Amsterdam, p. 693.<br />
WALLER, G. R. <strong>and</strong> FEATHER, M. S. (1983) The Maillard Reaction in Food <strong>and</strong> Nutrition. ACS Symposium<br />
Series No. 215. American Chemical Society, Washington DC.<br />
WHITTLE, N., ELDRIDGE, H., BARTLEY, J. <strong>and</strong> ORGAN, G. (1999) J. Inst. <strong>Brewing</strong>, 105, 89.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
10<br />
Wort boiling, clarification, cooling <strong>and</strong><br />
aeration<br />
10.1 Introduction<br />
Sweet wort is boiled with hops in acopper (kettle, hop-boiler). Sometimes, if the copper<br />
is not immediately available, the wort is held in an intermediate vessel, or underback,<br />
before it is boiled. It is held hot (75±80ëC; 167±176ëF) to minimize the risk of microbial<br />
infection.This`hold'shouldnotbeprolonged assomethermophilicorganisms, including<br />
some that can reduce nitrate ions to nitrite ions, can continue to grow <strong>and</strong> the wort will<br />
darken <strong>and</strong> its flavour will alter. Avariety of hop preparations may be used instead of<br />
whole hop cones (Chapter 7), necessitating the use of different types of equipment to<br />
clarify the wort after the boil. The changes that occur during the boil are complex<br />
(Chapter 9) but at the end of the boil spent hops (whole cones or fragments) <strong>and</strong> flocks of<br />
precipitated material, the hot break or trub, should be suspended in a perfectly clear, or<br />
`bright' wort, the colour of which will have increased during the boil. The trub <strong>and</strong> spent<br />
hops are separated from the wort, which is then cooled <strong>and</strong> aerated or oxygenated, usually<br />
while being transferred to a fermenter, where it is pitched (inoculated) with yeast.<br />
For many years the hop-boil, which usually lasted for 1.5±2 h but sometimes longer,<br />
was regarded as a simple process, the only variations being the duration of the boiling<br />
period, the choice of hops, the hopping rate <strong>and</strong> whether the hops were added at the start<br />
of the boil, in the middle or near the end (late hopping, when aroma hops are added).<br />
However, the need to reduce the cost of boiling has resulted in the testing of different<br />
technologies for saving energy <strong>and</strong> the difficulties encountered have emphasized the<br />
complexity of the boiling process <strong>and</strong> the necessity of balancing the changes that occur.<br />
There is a range of objectives to be met by boiling (Hough et al., 1982; Miedaner <strong>and</strong><br />
Narziss, 1986; Narziss, 1993). The first is to evaporate water <strong>and</strong> so concentrate the wort.<br />
Traditionally, an evaporation rate of 10% or more of the collected wort volume/h was<br />
usual. The cost of evaporating so much water is high, because the consumption of energy<br />
is high, <strong>and</strong> so it is now usual to minimize evaporation. The boil also removes unwanted<br />
volatile substances. Simply reducing the evaporation rate or shortening boiling times<br />
gives flavour <strong>and</strong> other problems, but newer designs of boilers reduce or eliminate this<br />
difficulty, by favouring the evaporation of the unwanted volatile substances, <strong>and</strong> boils of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
60±90min., with evaporation rates of as little as 4%/h can be used in appropriate<br />
equipment. The boil also sterilizes the wort, or at least destroys the `vegetative' forms of<br />
microbes probably in the first 10±15min. Spores may survive this process. From the boil<br />
onwards wort is h<strong>and</strong>led under nearly aseptic conditions.<br />
The changes that occur during boiling include the dispersion of the hop resins <strong>and</strong> oil,<br />
the isomerization of some of the -acids, the conversion of dimethyl sulphide precursor<br />
(SMM; DMSP; 4.157) to DMS (4.158), the formation of new flavour <strong>and</strong> aroma<br />
compounds, (largely through Maillard reactions, which also give rise to coloured<br />
compounds <strong>and</strong> so adarkening of the wort), the denaturation <strong>and</strong> inactivation of residual<br />
enzymes carried forward from the mash <strong>and</strong> the denaturation <strong>and</strong> coagulation of the<br />
proteins that, combined with polyphenols, form the trub (which may weigh 20±70gdry<br />
wt./hl Chapter 9). The more vigorous <strong>and</strong> prolonged the boil the more complete the<br />
removal of coagulable proteins <strong>and</strong> the more resistant the beer to the formation of nonbiological<br />
haze, but it will have less good foaming properties. The inactivation of the<br />
enzymes stabilizes the composition of the wort in one sense, but of course the high<br />
temperature maintains the on-going chemical reactions. The wort needs to be sufficiently<br />
agitated to cause the denatured proteins to coagulate <strong>and</strong> form flocks. This process may<br />
be assisted by the addition of copper finings, mainly the negatively charged, sulphated<br />
polysaccharide -carrageenan, `Irish moss', added at 4±8 g/hl (Chapter 9). Instead some<br />
breweries use silica gel.<br />
While vigorous mixing is essential, shear forces must be minimized to prevent the<br />
flocks being disrupted. Local `overheating' at a heating surface should be avoided as the<br />
level of coagulable protein is unduly reduced <strong>and</strong> it can remove proteins that stabilize<br />
foam. Local overheating, as was common with older, direct-fired coppers, can also cause<br />
`burn-on' <strong>and</strong> caramelization of wort sugars <strong>and</strong> copper adjuncts, which gives rise to<br />
(usually unwanted) flavours <strong>and</strong> cleaning difficulties with the copper. The boil is also<br />
used to `refine' the flavour of the beer by evaporating unwanted volatile flavour <strong>and</strong><br />
aroma compounds. These are more volatile that water <strong>and</strong> so, under the correct<br />
conditions, they can be evaporated to the desired extent at relatively low water<br />
evaporation rates. In the past it was often assumed that, during boiling, some oxidation of<br />
the wort by entrained air was desirable. The newer view is that this is not so, as oxidation<br />
darkens the wort, reduces flavour stability <strong>and</strong> favours haze formation in beers. The<br />
importance of oxidation varies with the type of beer being made.<br />
Other substances, besides hop preparations <strong>and</strong> copper finings, that may be added to<br />
wort in the copper, include mineral or biologically prepared acids (for pH adjustment),<br />
salts, tannins, malt extracts, sugars <strong>and</strong> syrups. The addition of acids, such as lactic acid,<br />
reduces the pH of the wort as does the addition of calcium salts. Calcium ions, displacing<br />
hydrogen ions from phosphates <strong>and</strong> other molecules, giving salts that precipitate during<br />
boiling, reducing the pH by 0.1±0.2 units. The reduction of wort pH during the boil<br />
reduces the colour, gives beer with a `cleaner' flavour, <strong>and</strong> a better `break' formation, but<br />
hop utilization is reduced unless pre-isomerized preparations are used. Where zincdeficiency<br />
may occur in the fermenters, small amounts of zinc chloride (0.1±0.2 mg/l)<br />
may be added to the wort. Gallotannins are sometimes added in calculated amounts to<br />
precipitate `haze-sensitive' proteins with the trub, so stabilizing the beer. Other<br />
substances tested include active carbon, nylon powder, PVPP powder, kieselguhr <strong>and</strong><br />
bentonite, but these are probably only rarely or never used in this way now.<br />
In small-scale brewing malt extract may be dissolved to form wort, which is boiled<br />
with hops in the same operation. In larger-scale brewing sugars <strong>and</strong> syrups (copper<br />
adjuncts derived from cereals or starch <strong>and</strong> so differing in whether or not they contain<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
nitrogenous <strong>and</strong>othersubstances(Chapter2))maybeaddedtoincreasetheconcentration<br />
of the wort <strong>and</strong>/or adjust its fermentability. The material must be added with care to<br />
ensure that it is dissolved <strong>and</strong> thoroughly dispersed in the wort. Failure can result in<br />
syrupy deposits settling onto the heating surfaces, where they are caramelized <strong>and</strong> burnt.<br />
Copper adjuncts are often used to increase the concentration of worts for use in highgravity<br />
brewing. For example if one part of asyrup, SG 1150 (36.4ëPlato) is added to<br />
nineparts ofawortofSG1040(10ëPlato) the final, mixed wort willhave anSGofabout<br />
1051 (12.6ëPlato). Hop boiling is normally abatch process. Attempts have been made to<br />
develop continuous hop-boiling to allow all stages of beer production to run<br />
continuously, but only with limited success (e.g. Hall <strong>and</strong> Fricker, 1966; Hough et al.,<br />
1982; see below).<br />
10.2 The principles of heating wort<br />
Early wort boilers were made of cast iron, but by the early 20th century most were made<br />
of copper. While copper has advantages (Table 10.1), more recently boilers have been<br />
made of stainless steel, but the vessels are still often called `coppers'. Copper is easily<br />
worked, has ahigh thermal conductivity (approx. 380W/mKat 100ëC; 212ëF) <strong>and</strong><br />
vessels made from it have an attractive appearance (Hancock <strong>and</strong> Andrews, 1996;<br />
Royston, 1971; Wilkinson, 1991a). It can catalyse oxidative reactions, it increases wort<br />
colour during boiling <strong>and</strong> is said to remove sulphur-containing substances from wort,<br />
(perhaps by catalysing the oxidation of thiol groups), <strong>and</strong> so improves beer flavour. It<br />
also favours nucleate boiling because its surface is relatively wettable. Stainless steel,<br />
generally an austenitic grade such as 304 or 316, is less expensive but is not so easily<br />
worked, <strong>and</strong> so vessels are often of simpler shapes than those made of copper. It is more<br />
resistant to dilute acids or strongly caustic cleaning agents, such as 2±4% caustic soda,<br />
<strong>and</strong> because of its greater strength vessels made from it can be thinner (e.g. 1.6mm) than<br />
those from copper, so its lower thermal conductivity (approx. 167W/mKat 100ëC;<br />
212ëF),isnotveryimportant atheatexchangesurfaces.Thelesserwettabilityofstainless<br />
steel does not favour nucleated boiling or tolerance of high heat flux densities, so the<br />
maximum temperatures used at heat exchange surfaces are lower than those which may<br />
be used with copper.<br />
Earlycopperswere directly heated bywood orcoalfires,<strong>and</strong>hadcalorificefficiencies<br />
ofabout 40±50%. Solid fuelheating isnow rare,but some kettles are heated directly with<br />
gas or oil-fuelled flames, with calorific efficiencies of about 70%. The heated area must<br />
becovered bywortbefore heatisappliedtopreventcaramelization<strong>and</strong>charring.Thefire<br />
orflamesmustbeextinguishedbeforethewortiswithdrawn.Directheatingoftenimparts<br />
characteristic flavours to beers. As breweries became larger, some used many relatively<br />
Table 10.1 Some properties of copper <strong>and</strong> stainless steel (Hough et al., 1982)<br />
Property Copper Stainless steel<br />
Density (kg/m 3 ) 8930 7930<br />
Specific heat (J/kg K) 385 510<br />
Thermal conductivity (W/m K) 385 150<br />
Yield stress (MN/m 2 ) 75 230<br />
Heat flux (kW/m 2 ) 80* 60 y<br />
* For a conventional, jacketed kettle.<br />
y For flat, stainless steel panels.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
small coppers so that as wort was collected it could be boiled in the shortest possible<br />
time. This reduced the chances of microbial infection <strong>and</strong> the operation was flexible.<br />
When steam heating became established some breweries continued to use several small<br />
coppers toretaintheseadvantages<strong>and</strong>evenoutthedem<strong>and</strong>forsteam,whileothersbegan<br />
touseoneortwolargecoppers,whichhave lowercapital<strong>and</strong>maintenancecostsbuthave<br />
alarge steam dem<strong>and</strong> while the wort is being heated from its collection temperature to<br />
boiling. One advantage of pre-heating wort on its way to the copper is that this sudden<br />
steam dem<strong>and</strong> is reduced. Some breweries use hot water, often at 145±170ëC (293±<br />
338ëF), to heat the coppers. To keep the water liquid (prevent it boiling) it is held under<br />
substantialpressure (around 17bar).Thehot water systems, which are expensive, need to<br />
be well insulated <strong>and</strong> need substantial outgoing <strong>and</strong> return pipework between the heat<br />
sink <strong>and</strong> the furnace(s). Because the water contains asubstantial amount of heat sudden<br />
dem<strong>and</strong>s are more easily met than by steam. For heavily used, flat-sided coppers water<br />
heated systems have been preferred. Coppers have been boiled experimentally using<br />
microwave heating (Herrmann, 1999).<br />
Steam is now the most usual heating agent in breweries. Steam, raised in aboiler<br />
house, is used dry <strong>and</strong> saturated at arelatively low pressure, about 4bar, at about 148ëC<br />
(298ëF). (As the pressure is increased so is the boiling point of water <strong>and</strong> the temperature<br />
ofthesteam;AppendixA.11).At aheatexchangesurface thesteamcondenses, givingup<br />
its latent heat which, in acopper, heats the wort. Asudden large dem<strong>and</strong> for heat can<br />
cause the condensation of agreat deal of steam <strong>and</strong> so asudden drop in pressure which<br />
may cause vessels to collapse. Thus steam-heated systems must be fitted with pressure<br />
<strong>and</strong> vacuum release valves as well as condensate traps, pipes to return the condensed<br />
water to the boiler, strainers <strong>and</strong> pressure reducing valves. Delivering steam at alower<br />
pressure reduces the temperature of aheating surface. Steam is used in acopper with 90±<br />
95% efficiency, but steam raising is less efficient <strong>and</strong> the overall efficiency is 65±70%<br />
(Hough et al., 1982).<br />
In modern installations wort is nearly always heated while flowing upwards in pipes<br />
mounted in steam jackets. The conduction of heat under ideal conditions, at asteady state,<br />
isdescribedbyFourier'sLaw,qˆkA T=X,whereqistherateofheattransmission,kis<br />
the thermal conductivity of the material, Ais the cross-sectional area at right angles to the<br />
heat flow, Tis the temperature difference between the steam <strong>and</strong> the liquid to be heated<br />
<strong>and</strong> Xis the thickness of the material across which the heat is flowing. However, heat<br />
transfer from steam to wort is better described by amore empirical formulation because<br />
the system is quite complex <strong>and</strong> changes with time as scale <strong>and</strong> baked-on materials<br />
(fouling) accumulate on the heat transfer surfaces, impeding heat flow. As wort at boiling<br />
temperature flows into the base of aheating tube it flows in as asingle, liquid phase <strong>and</strong><br />
theflowshouldbefastenoughtobecometurbulent(Fig.10.1).Asitgainsheatitbeginsto<br />
boil <strong>and</strong> bubbles form, creating a second, vapour phase <strong>and</strong> reducing the density of the<br />
wort/steam bubble mixture in the tube. At first the bubbles form on <strong>and</strong> separate from the<br />
wall of the tube, giving saturated, nucleate boiling. However, if the temperature of the<br />
walls is too high the liquid may become separated from the wall by a film of steam. This<br />
film boiling is undesirable as heat transfer is impaired <strong>and</strong> solids can be deposited <strong>and</strong><br />
baked onto the heating surface, causing fouling. To avoid film boiling the temperature of<br />
the steam is limited by limiting the pressure to about 3 bar with stainless steel <strong>and</strong> 5 bar<br />
with copper. This difference is because the greater surface wettability of the copper<br />
favours the separation of steam bubbles from the heated surface <strong>and</strong> so favours nucleated<br />
boiling (Andrews, 1992; Hancock <strong>and</strong> Andrews, 1996).<br />
The transfer of heat from steam to wort is less simple than might be supposed. The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Rising film<br />
Saturated<br />
nucleate<br />
boiling;<br />
turbulent,<br />
two-phase<br />
flow<br />
Submerged<br />
boiling<br />
Turbulent,<br />
single-phase<br />
flow<br />
Two-phase flow<br />
Single-phase flow<br />
Annular flow,<br />
with entrainment<br />
Churn flow<br />
Slug flow<br />
Bubbly flow<br />
Tube wall<br />
Steam or hot<br />
water<br />
Fig. 10.1 Diagram of the stages of boiling in wort rising up a steam- or water-heated tube (after<br />
Wilkinson, 1991a, b). As the wort rises so boiling becomes more vigorous, the flow becomes `twophase'<br />
as steam vapour bubbles appear <strong>and</strong> come to occupy an increasing proportion of the volume.<br />
temperature in the steam, at a particular pressure, is supported by convection. However,<br />
the surface of the wall is covered by a laminar film of condensed water <strong>and</strong> this, in turn,<br />
may be separated from the metal wall of the heat exchanger by a layer of scale deposited<br />
from the steam. On the wort side of the wall there may be a deposit of baked-on organic<br />
material <strong>and</strong> outside this fouling, between it <strong>and</strong> the moving wort, there is a laminar,<br />
stationary wort film. The stationary film of condensate, the scale, the metal wall, the<br />
fouling <strong>and</strong> the stationary layer of wort must all conduct heat <strong>and</strong> all provide resistance to<br />
heat flow from the steam to the wort. Heat distribution in the wort flowing up the tube<br />
occurs by forced convection. To obtain good mixing <strong>and</strong> heat transfer to all the liquid the<br />
wort must have a turbulent flow, which will partly disrupt the stationary layer. With<br />
turbulent flow heat transfer in a liquid is proportional to the (velocity of flow) 0.8 . With the<br />
passage of time the scale <strong>and</strong> the fouling increase <strong>and</strong> so, in order to maintain wort<br />
boiling, the temperature of the steam side must be increased by increasing the steam<br />
pressure.<br />
The heat flow in a wort heater, Q, can be quantified as Q ˆ U A T, where U is the<br />
overall heat transfer coefficient of the system, A is the area at right-angles to the heat flow<br />
<strong>and</strong> T is the temperature difference between the bulk of the steam <strong>and</strong> the body of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 10.2 Typical heat transfer coefficients in wort heaters (kW/m 2 K) (Hough et al., 1982)<br />
Steam side (low pressure) 5±6<br />
High pressure, hot water side 2.5<br />
Wort side (clean) 1.6<br />
Wort side (dirty) 1.2<br />
Vessel wall (copper; k/x) 19.3<br />
Vessel wall (stainless steel; k/x) 7.5<br />
Overall heat transfer coefficient for a clean, stainless steel vessel 1.0<br />
As above, but for a dirty kettle 0.5<br />
wort. UA is called the heat flux, q, <strong>and</strong> is given in kW/m 2 . The upper limit of the heat<br />
flux, using stainless steel, is about 95, but in <strong>practice</strong> this is limited to 50 to reduce the<br />
risks of fouling. The steam condensate layer, the steam-side scale, the metal wall, the<br />
wort-side fouling <strong>and</strong> the wort stationary layer all contribute to U. With increased fouling<br />
or scaling U decreases. Thus this factor in the equation is empirical, it changes with time<br />
<strong>and</strong> must be determined by experiment (Table 10.2). In practical calculations allowance<br />
must be made for the fact that heat transfer is into a tube <strong>and</strong> not across a plane surface.<br />
At some point the heat exchanger must be cleaned to remove the scale <strong>and</strong> the fouling,<br />
the heat transfer coefficient, U, is increased <strong>and</strong> so processing can be resumed using<br />
lower pressure steam.<br />
Boiling is accompanied by evaporation <strong>and</strong> hence an increasing concentration of the<br />
wort. One ton of steam can evaporate about 15 m 3 of water from wort. The evaporation<br />
rate is monitored <strong>and</strong> as it tends to fall in successive brews, due to the fouling <strong>and</strong> scaling<br />
in the heat exchanger, the steam pressure/temperature is increased to maintain the<br />
evaporation rate. By indirectly monitoring the rate of fouling cleaning can be timed to be<br />
carried out when it is really needed. The evaporation rate may be monitored by<br />
· automatically determining the level of the boiling wort from the falling pressure at the<br />
base of the vessel, using a pressure transducer, or by measuring the depth of the liquid<br />
· measuring steam utilization or condensate production<br />
· automatic sampling <strong>and</strong> determination of the wort density (SG)<br />
· determining the density indirectly, in the base of the vessel at the inlet to an external<br />
cal<strong>and</strong>ria heater, by measuring the velocity of ultrasound generated by a transducer.<br />
This device is said to be accurate to within +/ 0.1% SG (Forrest et al., 1993).<br />
With older patterns of coppers cleaning was carried out every 6±12 brews, when the<br />
heat transfer rate might have fallen by as much as 25%. Reasons for infrequent cleaning<br />
included (i) the difficulty of finding time in a busy brewing schedule; (ii) the costs of<br />
cleaning including the costs of detergents, hot water, fitting <strong>and</strong> dismantling the<br />
equipment <strong>and</strong> the disposal of the effluent; (iii) oversizing the heat exchange surface <strong>and</strong><br />
(iv) the reduced need for evaporation as weak worts were recycled to the following brew.<br />
In newer external wort heaters, (EWH), with rapid, turbulent flow through the heatexchange<br />
tubes, some scouring of the tube surfaces occurs <strong>and</strong> the large heating areas<br />
allow the use of lower wall temperatures <strong>and</strong> so cleanings can be less frequent, for<br />
example having 30 brews between cleanings.<br />
If the pressure on wort is increased the boiling point rises, trub formation <strong>and</strong> the<br />
isomerization of -acids are accelerated, as are colour formation, DMS formation <strong>and</strong><br />
other changes. Thus by boiling at elevated temperatures boiling times can be reduced <strong>and</strong><br />
energy (steam consumption) can be saved. On the other h<strong>and</strong> the increased removal of<br />
protein results in beers with poorer foaming characteristics. It is said that for each 4 ëC<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(7.8ëF) rise in temperature the boiling time can be halved (Chapter 9). However, as the<br />
different processes in the wort have different temperature coefficients, the changes in the<br />
wort will alter at different relative rates at different temperatures <strong>and</strong> so the nature of the<br />
wort will alter. Some problems encountered in boiling at elevated temperatures are<br />
discussed later. Generally boiling wort at, or near, normal atmospheric pressure gives the<br />
most acceptable results.<br />
Steam-heated vessels may have jackets or welded semicircular pipes to carry the<br />
steam (Fig. 15.2). Sudden steam dem<strong>and</strong>s can induce adrop in pressure that can cause<br />
jackets to collapse. Pipe heating is more resistant to this kind of damage. Steam is<br />
distributed around a brewery in well-insulated mains at a pressure above that needed at<br />
the various sites. Valves reduce the pressure at each site to an appropriate extent to<br />
achieve the desired temperature. As heat is given up the steam condenses <strong>and</strong> the<br />
condensate is collected <strong>and</strong> returned to the boiler or sent to drain, without allowing steam<br />
to escape. Each heating zone must have a pressure gauge <strong>and</strong> a thermometer <strong>and</strong>, for<br />
safety's sake, a vacuum valve <strong>and</strong> a pressure release valve. In addition there must be<br />
bleed valves to allow the escape of air when the system is first filled with steam (Kunze,<br />
1996).<br />
10.3 Types of coppers<br />
The increasing sizes of brewing vessels <strong>and</strong> the need to reduce costs (by increasing<br />
heating efficiency) <strong>and</strong> to reduce the emission of vapours (to meet anti-pollution<br />
legislation) have led to the development of a variety of new types of coppers (hop-boilers,<br />
kettles). In several cases these, while technically successful in saving energy, produced<br />
worts with unacceptable or unusual characteristics. In consequence the use of some<br />
vessels has been discontinued while in other cases coppers found unsuitable by some<br />
brewers have been retained in use by others, perhaps because of the different<br />
characteristics of the beers being produced (Andrews, 1992; Andrews <strong>and</strong> Axcell<br />
(private communication); Clarke <strong>and</strong> Kerr, 1991; Hackensellner, 1999; Herrmann,<br />
1998a,b; Hind, 1940; Kunze, 1996; Miedaner, 1986; Narziss, 1986a, 1992, 1993;<br />
Ormrod, 1986; Rehberger <strong>and</strong> Luther, 1994; Schwill-Miedaner <strong>and</strong> Miedaner, 2002;<br />
Vermeylen, 1962; Wilkinson, 1985, 1991a, b).<br />
The oldest coppers were made of iron with cylindrical sides <strong>and</strong> rounded bases <strong>and</strong><br />
were open to the atmosphere. Often these were replaced by copper vessels of the same<br />
type. The copper was mounted in a brick housing with a furnace for burning solid fuel at<br />
the base <strong>and</strong> a flue that wound round the side of the copper to a chimney. At least one<br />
such copper is still in use. The copper must be filled with wort before the furnace is fired<br />
<strong>and</strong> the fire must be drawn before the copper is emptied to prevent the heated area<br />
becoming too hot, causing wort to burn on. Sufficient space must remain above the fill<br />
level to contain the boiling, frothing wort. Such open coppers release steam into the<br />
surroundings, creating unpleasant working conditions, condensation <strong>and</strong> drip-back that<br />
leads to deterioration of the building <strong>and</strong> cleaning difficulties. Whole hop cones are<br />
added by h<strong>and</strong>, in weighed amounts. As kettles became larger <strong>and</strong> more complex in shape<br />
they were increasingly made from copper. Usually they were covered with a dome<br />
provided with a chimney to carry steam outside the building <strong>and</strong> an inspection <strong>and</strong> access<br />
opening, which could be closed with sliding doors. The base was usually hemispherical<br />
but sometimes it was domed upwards, to encourage better circulation of the wort. There<br />
might be a mechanical stirrer, driven by a shaft from above. Sometimes stirrers,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Steam<br />
Air escape<br />
Condenser<br />
Steps for<br />
cleaning<br />
copper<br />
Propeller<br />
Steam inlet<br />
Condenser<br />
pipe<br />
Motor<br />
Wort<br />
gauge<br />
Propeller<br />
drive<br />
Floor<br />
Floor<br />
Fig. 10.2 An old pattern of copper having a rounded base <strong>and</strong> heated by a symmetrical, external<br />
steam jacket (after Hind, 1940).<br />
`rummagers', carried a series of loops of chain that swept the bottom of the vessel,<br />
dispersing deposited materials. It was appreciated that a vigorous `rolling boil' was<br />
desirable <strong>and</strong> that there needed to be an evaporation rate of 8±10% or even 15% of the<br />
original wort volume/h. As the boil often lasted 2±2.5 h there was a substantial reduction<br />
in the volume of the wort <strong>and</strong> hence an increase in its concentration. While this allowed<br />
the use of large volumes of sparge liquor (<strong>and</strong> hence a good extract recovery from the<br />
mash) it was time consuming <strong>and</strong> costly because of the large amounts of fuel needed to<br />
generate the heat needed to evaporate the water. Direct heating with solid fuels, or oil, or<br />
gas has become unusual.<br />
As copper sizes increased it became apparent that simply heating the base, for example<br />
with steam (Fig. 10.2), was inadequate because the heating surface area to wort volume<br />
ratio decreased with increasing vessel size <strong>and</strong> it was undesirable to overheat the heating<br />
surface. Some brewers used a number of small coppers <strong>and</strong> filled them <strong>and</strong> brought them to<br />
the boil in sequence, saving time <strong>and</strong> avoiding having to accumulate the wort in a large<br />
copper or underback, with a consequent risk of microbial infection. A partial solution was<br />
to have the heating area of the copper divided into zones. As the first, lower zone was<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Floor<br />
Fascia panel<br />
Manhole <strong>and</strong> inspection<br />
port<br />
Valve for emptying<br />
copper<br />
Floor<br />
Steam jackets<br />
asymmetrically placed<br />
Heating surfaces<br />
Wort discharge<br />
Fig. 10.3 Acopper with arounded base <strong>and</strong> asymmetrically placed steam jackets<br />
(Hough et al., 1982).<br />
covered with wort the heating could be switched on. Wort continued to enter the copper<br />
<strong>and</strong> when the second <strong>and</strong> third zones were covered steam to these was also switched on, in<br />
turn. Asecond problem was that the even application of heat to the base of acopper with a<br />
rounded base gave apoor wort circulation <strong>and</strong> inadequate mixing. To overcome this some<br />
coppers were fitted with mechanically driven impellers (Fig. 10.2) while others were fitted<br />
with asymmetric heating panels (Fig. 10.3). The wort adjacent to the heating surface<br />
exp<strong>and</strong>ed <strong>and</strong> became less dense, consequently it was driven upwards by the cooler, more<br />
dense wort that took its place. With asymmetric heating astrong circulation, with good<br />
mixing <strong>and</strong> asteady upflow of wort across the heating surface, could be achieved.<br />
A so-called `high-efficiency' copper became popular in Europe, <strong>and</strong> provided a<br />
reference for performance for anumber of years (Fig. 10.4; Table, 10.3; Schwill-Miedaner<br />
<strong>and</strong>Miedaner,2002;Narziss, 1992).With thisdesignsteamheatingwasappliedtothebase<br />
of the copper <strong>and</strong> to the centrally placed, truncated cone, increasing the heating surface<br />
area/wort volume ratio <strong>and</strong> inducing aconvective upward flow of wort in the centre of the<br />
vessel <strong>and</strong> acompensatory flow down the outside. The copper was fitted with an agitator,<br />
which assisted the flow <strong>and</strong> mixing. Another approach was to supplement or replace the<br />
surface heating panels with internal heaters. Sometimes these were coils of tubes, but these<br />
did not encourage good circulation in the wort <strong>and</strong> were difficult to clean. Other patterns,<br />
such as `star heaters', which were mounted centrally inthe bases of vessels <strong>and</strong>encouraged<br />
an upward, convective flow of acolumn of wort, were better (Fig. 10.5).<br />
Adifferent approach was to make flat-sided kettles of stainless steel, which were<br />
rectangular in plan (Fig. 10.6). Experience showed that wort circulation <strong>and</strong> mixing was<br />
inadequate, despite the asymmetrical disposition of the heaters, <strong>and</strong> flow had to be<br />
assisted by impellers. The headspace was inadequate to accommodate the foam that was<br />
generated <strong>and</strong> the foam had to be broken by downwardly directed jets of air. This was<br />
undesirable both from the point of view of favouring undesirable oxidations in the wort<br />
<strong>and</strong> reducing the chances of effective heat recovery by diluting the steam <strong>and</strong> vapour<br />
from the boil with air (Section 10.7).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 10.3 The characteristics of some conventional coppers <strong>and</strong> some other boiling systems<br />
(data of Schwill-Miedaner <strong>and</strong> Miedaner, 2002)<br />
System of boiling Temperature* Boiling time Evaporation y<br />
(ëC) (ëF) (minutes) (%)<br />
`High performance' copper 100 212 120±150 12±16<br />
Internal/external boilers, with 102±103 215.6±217.4 60±80 about 8<br />
back pressure<br />
Low-pressure boiler 103±104 217.4±219.2 55±65 6±7<br />
Dynamic low pressure boiling 103±104 217.4±219.2 45±50 4.5±5<br />
High temperature wort boiling 130±140 266±284 2.5±3 6±8<br />
Thin film heating 100 212 35±40 4±4.7<br />
* Temperature at the exit from the heater.<br />
+ Percentage of the original volume collected.<br />
Stack<br />
Manhole <strong>and</strong> inspection port<br />
Valve for discharging wort<br />
Floor<br />
Agitator Steam jacket in dimple<br />
Agitator shaft<br />
Wort discharge<br />
Steam jacket on base of copper<br />
Fig. 10.4 A `high-efficiency' boiler having steam heating jackets on the base <strong>and</strong> on the sides of<br />
the central, truncated cone (Hough et al., 1982).<br />
Fig. 10.5 A `star' steam-heater, designed to have a large surface area while remaining compact,<br />
<strong>and</strong> be suitable for mounting in the base of a copper (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a) Entry for<br />
loading<br />
hops<br />
Heating<br />
surface<br />
Vapour<br />
stack<br />
Drive motor<br />
Inspection<br />
window<br />
Impeller Support<br />
Heating<br />
surfaces<br />
Heating<br />
surfaces<br />
Fig. 10.6 (a) Aflat-sided, stainless steel copper (rectangular in plan) with an asymmetric heating<br />
surface of semicircular tubes, <strong>and</strong> an impeller to assist mixing the wort. This design maintained the<br />
same size of end wall, but provided vessels of differing capacity by altering the length of the vessel<br />
(after Hough et al., 1982). (b) <strong>and</strong> (c) Cross-sections of alternative patterns of flat-sided kettles<br />
(after Rehberger <strong>and</strong> Luther, 1994).<br />
The increasing costs of fuels <strong>and</strong> the desire to reduce manpower <strong>and</strong> increase<br />
efficiency have driven the development of new wort boiling systems. Objectives, which<br />
are inter-connected, include the reduction in the use of primary energy, (involving the<br />
shortening of boiling times <strong>and</strong> reducing evaporation rates), the recovery <strong>and</strong> re-use of<br />
heat from the copper vapours (<strong>and</strong> the wort coolers), while avoiding the over-production<br />
of warm water, achieving the correct degrees of protein coagulation <strong>and</strong> removal of<br />
volatile substances, the avoidance of excessive colour generation or off-flavours <strong>and</strong> the<br />
maintenance or enhancement of the quality of the wort <strong>and</strong> the beer made from it. The<br />
equipment should be easy to clean <strong>and</strong> maintain <strong>and</strong> should not require cleaning too<br />
frequently. At least in large breweries cleaning should be fully automatic (CIP; cleaning<br />
in place) <strong>and</strong> not require the direct use of manpower.<br />
Newer coppers usually employ internal or external heaters in which the wort passes<br />
upwards through tubes surrounded by asteam-heated chamber. Figure 10.7 shows an<br />
internal heater in which the wort flows upwards through the heating unit into a<br />
constricting tube (sometimes called aVenturi tube), emerges above the level of the wort<br />
<strong>and</strong> strikes adeflector plate which directs it back as aspray onto the surface of the wort.<br />
In some instances two plates are used to direct the spray to the edge <strong>and</strong> to the midpoint<br />
of the radius of the copper. The sprays ensure agood circulation of the wort, they serve<br />
to `beat-back' foam <strong>and</strong>, by breaking the wort-stream into small droplets, they create a<br />
large surface area from which volatiles can evaporate into the vapour stream, which<br />
leaves the copper via the chimney. Being immersed in the wort internal, heaters are<br />
efficient but their size is restricted by the geometry of the vessel. This means that the<br />
heating surfaces must be heated to arelatively high temperature to obtain the necessary<br />
heat flux to obtain avigorous boil <strong>and</strong> this, in turn, leads to faster fouling <strong>and</strong> more<br />
frequent cleaning (sometimes as often as every six brews). By restricting the flow of<br />
wort in the Venturi tube asmall back-pressure can be achieved, raising the boiling point<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
(b)<br />
(c)
Tubes carrying<br />
wort through the<br />
steam chamber<br />
Inflow of wort<br />
Steam<br />
inlet<br />
Deflector cap<br />
Upflow<br />
of wort<br />
Wort<br />
discharge<br />
Spray head<br />
Base of<br />
copper<br />
Steam <strong>and</strong><br />
condensate outlet<br />
Wort outlet<br />
Fig. 10.7 An internal heater <strong>and</strong> `fountain' for coppers in which the wort moves up the steamheated<br />
tubes by convection, its flow is constricted, then it emerges from beneath the wort level <strong>and</strong><br />
strikes the deflector plate which directs it down onto the wort surface (Various sources).<br />
of the wort to 102±103ëC (215.6±217.4ëF). In part the size limitation for internal heaters<br />
may be offset by increasing the depth of the copper under the heater which, in<br />
consequence, can be relatively larger (Fig. 10.8).<br />
If the wort has not been pre-heated it arrives at the copper at mashing <strong>and</strong> sparging<br />
temperatures, about 75ëC (167ëF), <strong>and</strong> must be heated to boiling in the copper. With<br />
simple internal tubular heaters it is sometimes found that violent pulsations occur until<br />
the pre-heating period is complete (Stippler et al., 1997). This phenomenon resembles<br />
`boiling with bumping', familiar to chemists. The static, cool wort in the tubes is heated<br />
to boiling <strong>and</strong> then wort <strong>and</strong> vapour escape violently to be replaced with more cool wort.<br />
The process is repeated until the bulk of the wort is nearly boiling, when asteady stream,<br />
driven by convection, flows upwards through the heater. Mechanically driven impellers<br />
have been installed below some heaters to drive the wort upward through the heater<br />
during the heating phase <strong>and</strong> so avoid the pulsations. An alternative approach is to pump<br />
wortfromtwoseparatesitesatthebaseofthecopper,usingthecastingpump,<strong>and</strong>deliver<br />
it into the base of the internal heater, creating aforced upward flow over the heating<br />
surfaces (Hackensellner, 1999). When the wort is boiling steadily good mixing occurs.<br />
All efficient modern coppers are `closed', <strong>and</strong> so the steam is not diluted with air once<br />
boiling is established <strong>and</strong> has driven the air from the system, <strong>and</strong> so is available for heat<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ecovery. Internal heaters are less easily cleaned than external heaters. Boils of 60±<br />
80min., with total evaporations of around 8% are often suitable. External tube-<strong>and</strong>-shell,<br />
wort heaters (EWH), previously called `cal<strong>and</strong>ria' heaters are often preferred (Andrews,<br />
1992; Andrews <strong>and</strong> Axcell (private communication); Hancock <strong>and</strong> Andrews, 1996).<br />
Their size <strong>and</strong> geometry is not restricted by the dimensions of the kettle. Consequently<br />
the heat-exchange surface area can be relatively large, allowing lower wall temperatures,<br />
so less fouling is likely. External heaters are easily cleaned <strong>and</strong> may be retro-fitted to<br />
existingcoppers. Inanearly version, wort wasdrawn fromthe baseofthecopperthrough<br />
an axial flow pump <strong>and</strong> then upwards through atube-<strong>and</strong>-shell `cal<strong>and</strong>ria' steam heater<br />
<strong>and</strong> back into the vessel into the base of aVenturi `fountain' <strong>and</strong> was directed back onto<br />
the surface of the wort by aspreader (Fig. 10.9a). The axial flow pump remained running<br />
until the wort was boiling, then it was shut off <strong>and</strong> circulation in the system was<br />
maintained by the `thermosyphon' effect, driven by the reduced density of the wort<br />
boiling in the external heater being displaced upwards by the cooler, denser wort in the<br />
body of the vessel.<br />
The original heaters were comparatively short (1±1.2m; 3.28±3.94ft.) <strong>and</strong> had wide<br />
wort-carrying tubes (c. 75mm; 2.95in. diameter) to accommodate the passage of whole<br />
hops. If milled or milled <strong>and</strong> pelleted hops are used longer (2.5±4.5m; 8.2±14.76ft.) <strong>and</strong><br />
narrower tubes (25±65mm; 1.00±2.59in.) may be used having greater heat exchange<br />
efficiency. Avessel to boil 1000hl (611brl) of wort, <strong>and</strong> equipped with an external<br />
heater,might be5.8m(19.03ft.)indiameter <strong>and</strong>the straight sidesbe4m(13.12ft.)high.<br />
External wort heaters have been developed in two directions. In one the wort is pumped<br />
through the heater continuously during heating up <strong>and</strong> during the boil. In some cases<br />
these heaters may be mounted horizontally <strong>and</strong>/or the tubes may be bent into adistorted<br />
`S' shape, so that they traverse the heating steam chamber three times. At the exit of the<br />
heater arestriction valve may provide back-pressure raising the temperature of the wort<br />
Vent<br />
Inspection port<br />
Deflector cap<br />
Heater<br />
Sump<br />
Steam inlet,<br />
condensate<br />
outlet<br />
Fig. 10.8 Adiagram of acopper having an internal heater, mounted over <strong>and</strong> extending partly<br />
into, a `cup' or `sump'. The extra depth permits the heater to be larger than would otherwise be the<br />
case, <strong>and</strong> so it has a large heat transfer surface (after Michel, 1991).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 10.9 (a) A copper heated by a comparatively small external shell-<strong>and</strong>-tube cal<strong>and</strong>ria heater<br />
(Hough et al., 1982). This is an older design in which the pump initiated wort circulation but which<br />
relied on the circulation being driven by the thermosyphon effect once boiling was established. (b) In<br />
newer designs the wort flow is directed around the pump in a bypass loop when boiling is occurring.<br />
Wort from the EWH may be discharged above or at the wort surface. (Courtesy of Briggs of Burton,<br />
plc.). Alternatively, the wort may be discharged through a `fountain', as in the older pattern.<br />
in the heater to 103±104 ëC (217.4±219.2 ëF), allowing a shortened boil of 60±70 min.<br />
Although higher temperatures can be attained (e.g. 110 ëC; 230 ëF) they are generally not<br />
used because wort quality can be impaired (Narziss, 1993; Narziss et al., 1992). The<br />
release of pressure at the valve favours the rapid production of small vapour bubbles,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
whichareefficientatcarryingunwantedvolatilesawayintothevapourstream.Usingthis<br />
system the heater must be pressure resistant, the pump must run continually, with<br />
implications for maintenance <strong>and</strong> running costs, <strong>and</strong> the shear provided in the pump <strong>and</strong><br />
the restriction valve tends to break up the flocs of trub (hot break), complicating the<br />
subsequent clarification of the hot wort.<br />
Other external heaters (EWH) rely on the thermosyphon effect to drive the circulation<br />
of the wort once the system has reached boiling (Andrews, 1992; Andrews <strong>and</strong> Axcell<br />
(private communication); Hancock <strong>and</strong> Andrews, 1996; Wilkinson, 1991a,b; Fig. 10.9b).<br />
Whilethe wortis beingheatedtoboilingitisrecirculatedthroughtheEWH byapumpin<br />
abypass loop. Once boiling is established, <strong>and</strong> can be driven by the thermosyphon effect,<br />
the pump is bypassed, the wort flow is directed to asimple loop of obstruction-free<br />
pipework from the base of the copper into the base of the, vertically mounted, EWH. The<br />
absenceofpumpingsavespower,<strong>and</strong>wearonthepump<strong>and</strong>avoidstheliquidshearinthe<br />
pump. The large sizes of EWHs, that can be used, allow low heating temperatures to be<br />
applied <strong>and</strong> hence much slower fouling. For example, with 7% evaporation/h, aheat<br />
transfer of 4.63kW/hl <strong>and</strong> the same heat transfer coefficient then atypical internal heater<br />
might have aworking area of 0.08m 2 /hl while an external heater could have 0.20m 2 /hl.<br />
The steam/wort temperature differences would be 37ëC <strong>and</strong> 15ëC (66.5ëF <strong>and</strong> 27ëF)<br />
respectively (Andrews <strong>and</strong> Axcell, private communication). Wort passes through these<br />
heaters at the rate of 6±10 or even 12 vessel volumes/h. The scouring effect of the wort<br />
<strong>and</strong> the low wall temperatures needed to maintain the correct heat flux minimize fouling<br />
<strong>and</strong> so the necessary cleaning frequency may be less than once in 30 brews.<br />
The temperature of the wort in the heater (not in the body of the kettle) may reach<br />
105ëC(221ëF).Thewortemergesfromthereturnpipe<strong>and</strong>isinjectedtangentially,bythe<br />
force of the flow, either into the wort or, preferably, just above its surface. In each case<br />
the wort is driven to rotate in the kettle, giving good mixing, <strong>and</strong> in the second case the<br />
vigorous breakout of the vapour bubbles, initiated by nuclear boiling, efficiently<br />
evaporates a proportion of the volatiles. Evidently this arrangement is suitable for<br />
combined kettle/whirlpool vessels (Section 10.8). The high efficiency of volatiles<br />
removal, (only 4% evaporation may be needed), reduces the need for lengthy boiling <strong>and</strong><br />
soshorter,lessenergy-costlyboiling,ispossible.This,inturn,meanslessprimaryenergy<br />
is needed, less fuel is used <strong>and</strong> so carbon dioxide emissions are reduced. In afew small<br />
breweries plate <strong>and</strong> frame heat exchangers are used as EWHs.<br />
Arecent wort-boiling `Merlin' system uses athin film boiling <strong>and</strong> evaporation unit<br />
(Fig.10.10;Schwill-Miedaner<strong>and</strong>Miedaner,2002;Stippler,2000).Inthissystemwortis<br />
pumped from a whirlpool vessel (Section 10.8) <strong>and</strong> is discharged onto the apex of a metal<br />
cone, the upper third <strong>and</strong> lower two-thirds of which are independently heated by steam.<br />
The cone is enclosed in a vessel with a vapour-escape chimney. The wort flows<br />
downwards <strong>and</strong> boils in a thin, turbulent film over the heating surface <strong>and</strong> is then<br />
collected in a circumferential gulley from which it flows back to the whirlpool tank. This<br />
arrangement ensures a good evaporation of volatiles. The returning wort enters the tank at<br />
two locations, at the centre <strong>and</strong> tangentially at the periphery, where its entry drives the<br />
rotation of the vessel's contents. The two points of entry favour mixing. The heating<br />
phase, during which hops are added, may be for 35±40 min., with both zones of the<br />
heating cone at 130 ëC (266 ëF). The boil, which lasts about 40±60 min., takes place with<br />
only the lower zone being heated, to c. 120 ëC (248 ëF). During the heating-up period the<br />
wort is pumped at a rate of about 4±6 vessel volumes/h, while during the boil the rate is 4<br />
volumes/h. At the end of the boil the heater <strong>and</strong> pump are turned off <strong>and</strong> then, after the<br />
whirlpool rest (which may last as little as 10 min. because initial trub separation occurred<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Steam<br />
Valve<br />
Valve<br />
during the boil; Section 10.8), the clarified wort is pumped out over the heating cone to<br />
the cooler. At this stage only the lower part of the cone is heated to 116±120ëC (c. 241±<br />
248ëF). This `volatiles stripping' stage reduces the level of DMS very substantially. The<br />
total evaporation is about 4.5% of the initial wort volume, of which 1.5±2.5% is removed<br />
during the boil, the remainder during stripping. Thus this boiling system is energy<br />
efficient.<br />
10.4 The addition of hops<br />
Wort flow to top of cone<br />
Central<br />
wort entry<br />
Hop-dosing<br />
device<br />
Pump<br />
Pump<br />
Collection <strong>and</strong><br />
whirlpool tank<br />
Thin film evaporator/heater<br />
Spreading cone<br />
Upper heating zone<br />
Lower heating zone<br />
Wort collection gully<br />
Pump<br />
Tangential<br />
wort entry<br />
Water<br />
Heated<br />
water<br />
Cooled<br />
wort<br />
Fig. 10.10 Adiagram of a`Merlin' wort boiling system in which the wort is collected in a<br />
whirlpool vessel <strong>and</strong> is heated, boiled <strong>and</strong> volatiles are stripped by pumping it in a thin film over a<br />
steam-heated cone in a separate vessel (after various sources, including Anton Steinecker<br />
literature).<br />
In older <strong>and</strong>/or smaller breweries it is normal to add weighed amounts of hops (whole<br />
hops, milled hops, pelleted milled hops or hop extracts) to the coppers by h<strong>and</strong> (Chapter<br />
9). Traditionally, English beers were hopped at rates of 0.5±3.5lb./imp. brl (0.139±<br />
0.970kg/hl; Hind, 1940; Chapter 9). Manual additions are hazardous as the wort may<br />
suddenly boil up <strong>and</strong> over, <strong>and</strong> late additions, as with aroma hops, mean that air is<br />
admitted to the copper, which interferes with heat recovery from the vapour. There seem<br />
to be no good ways of automating additions of whole hops, but means of adding hop<br />
powders, pelleted hop powders <strong>and</strong> hop extracts have been automated (Anon., 1994;<br />
Benitez et al., 1997; Boyes, 1993; Kollnberger, 1986, 1987; Kunze, 1996; Langenhan,<br />
1995). Additions may be regulated by weight or by -acid content <strong>and</strong> several additions<br />
may be made at different stages of the boil. Hops may arrive at the brewery in bales,<br />
pockets, foil-lined bags or boxes. Extracts come in cans or in 80- or 200-litre mild steel<br />
disposable drums or in 1000-litre, returnable stainless steel drums.<br />
The hops or hop preparations must be stored cool <strong>and</strong> dry. Often the store is not<br />
adjacent to the copper(s). Thus the hops must be unpacked (containing any dust that is<br />
generated), be transferred to the brewhouse <strong>and</strong> weighed amounts be delivered to each<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ew at the correct stage. Unpacking can be laborious <strong>and</strong>, if automated, only one or two<br />
types of packing can be h<strong>and</strong>led conveniently. In one system boxes of pellets are<br />
conveyed forwardon aroller conveyor bygravity <strong>and</strong>, when the depth ofpelletsin afeed<br />
hopper has fallen to a predetermined extent, a box is opened, by two saws that<br />
automatically remove the ends. The empty box is discarded, its place being taken by the<br />
next full box, while the pellets fall into the hopper. Discharge from the hopper may be<br />
through a vibrating feeder into a weigher that determines the dose. Pellets can be<br />
transferred by mechanical or pneumatic means, provided these are not too damaging, or<br />
they can be broken up <strong>and</strong> be pumped after being slurried with wort or water. Pellets can<br />
become `sticky' <strong>and</strong> conveyers become contaminated <strong>and</strong> need regular cleaning. Pellets<br />
may beweighed into aconveyor<strong>and</strong> delivered directly into the copper, precautions being<br />
taken to prevent steam wetting the conveying system.<br />
With pressurized coppers the dose passes through the first valve, which is then closed,<br />
the pressure is equalized, then the second valve is opened <strong>and</strong> the hops are discharged<br />
into the vessel. In adifferent system each batch of hops or hop extract is weighed into a<br />
smallpressurevessel<strong>and</strong>,attheappropriatetimethehop`dose'isflushedintothecopper<br />
with hot water or hot wort. If different types of hops are to be added at different stages of<br />
the boil each batch must be in adifferent vessel. In yet another system hop pellets or<br />
extracts are slurried in hot water or wort <strong>and</strong> aliquots are pumped into coppers as<br />
required. Hop extracts may be added by suspending punctured tins in baskets immersed<br />
in the boiling wort. Alternatively, drums of extract may be pre-heated to not more than<br />
45±50ëC (113±122ëF) for 24h to reduce the viscosity, then the extract can be added to<br />
thewortflowingintothecopperviaameteringpump,ormaybeaddedwithaflushofhot<br />
wort or water. Hop oils may be added in at the end of the boil <strong>and</strong> it has been proposed<br />
that hop oils in the vapour condensate could be collected <strong>and</strong> added back in the same<br />
way.<br />
10.5 Pressurized hop-boiling systems<br />
The boiling point of water or wort depends on the pressure, (Appendix A.11), <strong>and</strong> so<br />
varies with changes in the atmospheric pressure. In breweries situated at high altitudes it<br />
is necessary to be able to seal <strong>and</strong> pressurize the coppers to elevate the boiling<br />
temperature to at least about 100ëC (212ëF). It has long been realized that by boiling<br />
wort at elevated temperatures hop -acid utilization is increased <strong>and</strong> potentially wort<br />
boiling can be shortened with savings in time <strong>and</strong> energy (Chapter 9). The acceptability<br />
of this approach is disputed. Recently, three types of elevated temperature/pressure<br />
boiling systems have been advocated, mainly in continental Europe. These have been<br />
called `low-pressure boiling', `dynamic low-pressure boiling' <strong>and</strong> `continuous, highpressure<br />
boiling'.<br />
10.5.1 Low-pressure boiling<br />
Low-pressure boiling involves increasing the `overpressure' in the copper (the pressure<br />
above atmospheric) to a relatively small extent, giving absolute pressures of up to 2 bar.<br />
This can be carried out in sealed <strong>and</strong> strengthened coppers with internal or external<br />
heaters. Temperatures of up to 110 ëC (230 ëF) have been used, but lower temperatures<br />
seem to be preferred, at pressures of 1.5±1.9 bar (Herrmann, 1985; Narziss, 1986a, b). A<br />
problem with low-pressure boiling is an inadequate evaporation of volatiles from the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
wort. In aparticular schedule wort is boiled for ten minutes at atmospheric pressure,<br />
when 1.3% evaporation occurs. The copper is sealed <strong>and</strong> the pressure rises, over ten<br />
minutes, until 106ëC (222.8ëF) is reached <strong>and</strong> 0.5% evaporation has occurred. This<br />
temperature is maintained, during boiling, for 18 minutes, when a further 1.3%<br />
evaporation takes place. Over 15 minutes the pressure is reduced until the temperature<br />
has declined to 100ëC (212ëF), allowing evaporation of 1.7%. Finally, the wort is boiled<br />
at atmospheric pressure for ten minutes, achieving afurther 1.3% evaporation, giving<br />
6.1% (6±7%) evaporation in total. This reduces boiling time by 20±30%, relative to<br />
unpressurized controls, <strong>and</strong> is convenient for energy recovery from the vapour (Section<br />
10.7).<br />
10.5.2 Dynamic, low-pressure boiling<br />
Dynamic, low-pressure boiling is designed to achieve a more rapid evaporation of<br />
volatiles <strong>and</strong> so asmaller, less costly total evaporation rate. Where the evaporation in a<br />
`traditional' boil is 8±10%/h, <strong>and</strong> in alow-pressure boil is 6±7% in total, that achieved in<br />
adynamic, low-pressure boil may be 4±5% in total (Kantelberg et al., 2000). In this<br />
system the pressure in the copper is allowed to increase, e.g., to 1.17 bar, <strong>and</strong> then is<br />
allowed tofall,e.g.,to1.05bar,with 6±7pressurechanges/h,givingtemperaturechanges<br />
between 101±102 <strong>and</strong> 104±105ëC (213.8±215.6 <strong>and</strong> 219.2±221ëF). Each time the<br />
pressure is released there is amassive release of small bubbles throughout the wort,<br />
which carry volatiles to the surface. The reduction in unwanted volatiles is good, with a<br />
total evaporation of 4.5±4.8%.<br />
10.5.3 Continuous, high-pressure boiling<br />
Continuous,high-temperaturewortboilinghasbeentriedinvarioussystems,withlimited<br />
success (Chantrell, 1983, 1984; Grasman <strong>and</strong> van Eerde, 1986; Rehberger <strong>and</strong> Luther,<br />
1994). The most successful attempts involved heating at 130±140ëC (266±284ëF) for<br />
150±180s. Savings in steam utilization of 60±65% were claimed (Fig. 10.11). Wort from<br />
a collection vessel at around 72 ëC (161.6 ëF), to which hop extracts are added, is<br />
successively heated in three heat exchangers to 90 ëC (194 ëF), 106 ëC (222.8 ëF) <strong>and</strong><br />
140 ëC (284 ëF) using vapour from the expansion tanks in the first two heaters <strong>and</strong> live<br />
steam in the third as the heating agents. The wort is maintained for three minutes at<br />
140 ëC (284 ëF) in a holding tube, then it is cooled in two stages in expansion vessels<br />
which successively reduce the temperature to 120 ëC (248 ëF) <strong>and</strong> 100 ëC (212 ëF). The<br />
vapours, which flash off during boiling only in the flash-tanks, remove volatiles <strong>and</strong><br />
transfer their heat to the incoming wort stream through the first two heat exchangers. The<br />
wort is then sent to the whirlpool/clarifier. The system is energy-efficient, but wort is<br />
darkened <strong>and</strong> good beer qualities have not always been achieved. The fouling of the<br />
equipment at the high temperatures used necessitates frequent cleaning using strong<br />
caustic solutions <strong>and</strong> hydrogen peroxide.<br />
10.6 The control of volatile substances in wort<br />
It is apparent, from the points made above, that the removal of unwanted flavour <strong>and</strong><br />
aroma volatiles is essential <strong>and</strong> that this may be achieved, in the appropriate equipment,<br />
without the traditional high evaporation rates. In many cases the removal of volatiles<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Heat<br />
exchangers<br />
From wort<br />
collection tanks<br />
Condensate<br />
Vapour, 100°C<br />
Vapour, 110°C<br />
Live steam<br />
6 bar<br />
90°C 106°C<br />
HE 1 HE 2 HE 3<br />
Condensate<br />
140°C<br />
Holding<br />
tubes<br />
120°C<br />
Expansion,<br />
evaporation<br />
vessels<br />
achieved during the copper-boil is not sufficient because, as with DMS, more are formed<br />
during the whirlpool rest or during the hold in the settling tank, periods during which the<br />
hot wort is clarified. Boiling wort most readily gives up volatiles when it is finely<br />
dispersedasinacopperwithaVenturitube<strong>and</strong>spreaderorwhenitisdischargedfroman<br />
external heater in atwo-phase mixture of liquid <strong>and</strong> vapour bubbles above the surface, or<br />
if the pressure on the boil is reduced causing the creation <strong>and</strong> escape of many small<br />
bubbles. Evaporationfrom coppershas alsobeen improved,<strong>and</strong>the boilingtimereduced,<br />
at least experimentally, by sparging the boiling wort with an inert gas, nitrogen,<br />
introduced at the base of the copper (Mitani et al., 1999). Maule <strong>and</strong> Clark (1985)<br />
demonstrated that boiling might be avoided by simmering the wort at 93ëC (199.4ëF),<br />
removing sufficient protein by treatment with silica hydrogel <strong>and</strong> removing volatiles by<br />
spraying the wort into an evaporation chamber held under apartial vacuum. Volatiles<br />
removal also occurs in the flash evaporation chambers used in continuous, hightemperature<br />
boiling (Fig. 10.11), <strong>and</strong> is of value in treating worts being transferred from<br />
the clarification vessel to the cooler (Lustig et al., 1997).<br />
Other approaches have been used to remove unwanted volatiles, generated in the hot<br />
wortduringclarificationrests.Thiscanbeachievedbyheatingathinlayerofwortduring<br />
the transfer from the whirlpool tank to the cooler (Stippler, 2000; Fig. 10.10). An<br />
alternative approach is to `steam-strip' the wort (Bonachelli et al., 2001; Braekeleirs <strong>and</strong><br />
Bauduin, 2001; Seldeslachts et al., 1997; Fig. 10.12). In the stripper the wort is preheated<br />
to boiling <strong>and</strong> is then sprayed into the top of acolumn packed with rings <strong>and</strong><br />
saddles giving ahigh surface area of 50±500m 2 /m 3 .As the wort percolates downwards,<br />
asathinfilmoverthecolumnpacking,itmeetsaslowcounter-flowoflivesteammoving<br />
upwards. This adjustable process carries away volatiles to acondenser <strong>and</strong> 0.5±2%<br />
evaporation occurs. Heat may be recovered, as hot water, from the condenser. The<br />
stripped wort moves directly to the cooler.<br />
By cooling the wort, e.g., to 89ëC (192.2ëF), during its transfer from the copper to the<br />
whirlpool tank, the reactions proceeding in the wort in the whirlpool are usefully slowed<br />
EV<br />
1<br />
EV<br />
2<br />
100°C<br />
Wort to whirlpool<br />
separator<br />
Fig.10.11 Aschemeofahigh-temperaturewortboilingsystem(afterChantrell,1983;Clarke<strong>and</strong><br />
Kerr, 1991). The vapours from the expansion vessels (EV 1 <strong>and</strong> 2) heat the wort in the first two heat<br />
exchangers (HE 1 <strong>and</strong> 2). Heat is provided to the third heat exchanger, (HE 3), raising the<br />
temperature of the wort to 140 ëC (284 ëF), by live steam. This temperature is held for three min.<br />
during the passage of the wort through the holding tube. The reduction in temperature <strong>and</strong><br />
evaporation occurs in two controlled steps, when the wort is boiled in the expansion vessels <strong>and</strong><br />
unwanted volatiles are removed.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Steam<br />
Wort preheater<br />
Column packing<br />
Condensate<br />
Wort<br />
Wort receiver<br />
Wort distribution<br />
system<br />
<strong>and</strong> fewer volatiles are formed. Cooling may be achieved using cold water <strong>and</strong> a plate<br />
heat exchanger (producing warm water), or by expansion in a vacuum chamber (Lustig et<br />
al., 1997; Krottenthaler <strong>and</strong> Back, 2001; Krottenthaler et al., 2001).<br />
10.7 Energy conservation <strong>and</strong> the hop-boil<br />
Heated water<br />
Vapour<br />
condenser<br />
Supporting grid<br />
Vapour condensate<br />
Cooling water<br />
Steam to injector<br />
Stripped wort<br />
Fig. 10.12 A steam stripper for removing volatiles from wort (after Bonachelli et al., 2001;<br />
Seldeslachts et al., 1997).<br />
Energy <strong>and</strong> warm water production <strong>and</strong> consumption link all parts of the brewing process<br />
(Manger, 1998; Schu, 1995; Unterstein, 1992). As far as possible `waste' heat, that is heat<br />
other than that derived directly from primary heating, must be conserved <strong>and</strong> utilized.<br />
This implies `good housekeeping', well-insulated vessels, no steam or other leaks,<br />
automatic controls where possible, efficient designs of plant <strong>and</strong> brewing operations, <strong>and</strong><br />
well-trained staff. Thus, besides steam <strong>and</strong> other `direct' heating, heat as vapour or hot<br />
water is produced in mash- (decoction) <strong>and</strong> adjunct-cooking, in wort boiling <strong>and</strong> in wort<br />
cooling. Hot water can be or is used in pasteurizers, in pre-warming lauter tuns or other<br />
vessels, in mashing in, in sparging, in preheating wort before it reaches the copper (to<br />
avoid a sudden high dem<strong>and</strong> for steam), in copper boiling, in CIP <strong>and</strong> other plant cleaning<br />
<strong>and</strong> in cleaning cans, kegs, casks <strong>and</strong> bottles as well as space heating, in preheating boiler<br />
feed <strong>and</strong> even heating anaerobic waste digestion plant. This section is particularly<br />
concerned with saving heat during hop-boiling, but it is unrealistic to consider this in<br />
isolation from the rest of the brewing process (Boer, de 1991; Clarke <strong>and</strong> Kerr, 1991;<br />
Fohr <strong>and</strong> Meyer-Pittroff, 1998; Herrmann, 1998a, b; Kunze, 1996; Lenz et al., 1991;<br />
Miedaner, 1986; Narziss, 1993; Reed, 1992). Heat recovery from boiler vapours requires<br />
that the system is purged <strong>and</strong> air-free. The <strong>practice</strong> of letting air into the copper, to reduce<br />
fobbing <strong>and</strong> to increase the draught up the stack to encourage evaporation, as well as<br />
encouraging wort oxidation, blocks heat recovery (Hough et al., 1982).<br />
Heat recovery from the vapours generated during hop-boiling has attracted much<br />
attention. In one, relatively simple, system vapour from pairs of boiling coppers was fed<br />
into a main <strong>and</strong> was condensed initially with jet condensers but subsequently, <strong>and</strong> with<br />
much greater efficiency, by cold or warm water sprays (Morris, 1987). A rise in<br />
temperature in the main, caused by the arrival of vapour from the boiling coppers,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
automatically triggered the first of a pair of sprays to come on. If the temperature of the<br />
main downstream from the first spray rose, indicating incomplete vapour condensation, a<br />
second, back-up spray came into operation. A key factor was the use of decarbonated<br />
water in the sprays, to avoid the deposition of scale in the pipes. Heat recovery in the<br />
resultant hot water was > 95%. Cold water at about 7 ëC (45 ëF) was used initially in the<br />
sprays <strong>and</strong> water with vapour condensate at about 90 ëC (195 ëF) was produced. Later<br />
warm water was used in the sprays. The small amounts of organic materials present in the<br />
hot water had no unwanted effects. Only a small amount of cooled vapour was discharged<br />
outside the building, <strong>and</strong> so the escape of organic materials <strong>and</strong> odours was negligible.<br />
Vapour condensers, which are widely used, produce hot water by heating cooling<br />
water, while the vapours from boiling wort (or mash) are condensed. The hot condensate<br />
may be cooled further in a heat exchanger before being used as a first cleaning rinse or<br />
being directed to waste. The warmed water produced when cooling the condensate may<br />
be used in various ways, including being the cooling feed to the vapour condenser, where<br />
it is heated to a substantially higher temperature. Copper condensers, or economizers, are<br />
not new; one was illustrated in 1875 (Narziss, 1986b). They may be used with internally<br />
or externally heated coppers, boiling at either ambient or elevated pressures. They are<br />
invariably used in thermal vapour recompression systems (see below). A system in which<br />
hot water from a condenser is used, in conjunction with a hot water, `energy storage tank'<br />
to heat a mash-mixing vessel is shown in Fig. 10.13. The temperature of the heated water<br />
is regulated by how the system is operated but is generally between 80 <strong>and</strong> 98 ëC (174.2<br />
<strong>and</strong> 208.4 ëF). In principle, it is desirable to use `waste' heat as it becomes available.<br />
However, as brewing schedules often cannot be arranged to achieve this it is necessary to<br />
have well-insulated hot water storage tanks to act as `energy stores'. These are arranged<br />
in various ways, for example, single, tall <strong>and</strong> relatively narrow tanks with thermal<br />
gradients (Fig. 10.13) or linked pairs of tanks may be used (Vollhals, 1994).<br />
Wort collection<br />
vessel<br />
Mash mixing<br />
vessel<br />
72°C<br />
From<br />
lauter<br />
tun<br />
Wort<br />
preheater<br />
95°C<br />
Heated<br />
water<br />
Cool<br />
water<br />
99°C ~80°C 78°C 99°C<br />
To whirlpool<br />
Kettle vapour<br />
condenser<br />
78°C<br />
Condensate<br />
cooler<br />
Waste<br />
Deflector<br />
Internal<br />
boiler<br />
99°C<br />
99°C<br />
78°C<br />
Energy<br />
storage<br />
tank<br />
Fig. 10.13 An arrangement where vapour from a kettle is condensed <strong>and</strong> heat recovered in hot<br />
water is stored in a temperature gradient tank <strong>and</strong> is partly used to heat the wall jacket of a mash<br />
mixing vessel <strong>and</strong> the wort from the lauter tun, before it enters the copper (after Herrmann, 1998b).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Valve<br />
Vapour<br />
Copper<br />
Vapour, 100°C<br />
Wort, 106°C<br />
Wort<br />
spreader<br />
Wort return, 100°C Pump<br />
Mechanical<br />
vapour compressor<br />
Compressed<br />
vapour, 120°C<br />
Live steam supply<br />
Cal<strong>and</strong>ria reactor<br />
Condensate<br />
Heated water, 85°C<br />
Pump<br />
Cooling water, 15°C<br />
Condensate<br />
cooler Cooled<br />
condensate,<br />
30°C, to waste<br />
Fig. 10.14 Outline of amechanical vapour recompression (MVR) system (after Fortuin, 1995).<br />
During boiling the wort, passing through the external wort heater, is heated by the compressed<br />
vapours from the copper, supplemented with live steam as needed. Condensate from the wort heater<br />
is cooled, water being heated in the process.<br />
Wort cannot be boiled with water vapour at 100ëC (212ëF). However, if the vapour is<br />
compresseditstemperaturerises<strong>and</strong>,provided thattheheatexchange areaisadequate,as<br />
can most easily be arranged with external copper heaters, boiling can be supported by the<br />
heat in the compressed vapour. In amechanical vapour recompression system (MVR),<br />
once all the air has been displaced from aclosed copper, acompressor (screw-, turbo- or<br />
rotary piston-compressor; often aRootes compressor) is switched on. The vapour is<br />
compressed to have an overpressure of 0.2±0.7 bar, <strong>and</strong> atemperature of 108±112ëC<br />
(226.4±233.6ëF). This vapour is returned to the copper heater <strong>and</strong>, by condensing <strong>and</strong><br />
giving up its latent heat in the normal way, maintains boiling, with supplementary<br />
additions of live steam if necessary (Fig. 10.14). This system uses electrical energy to<br />
drive the compressor. Compressors can be expensive <strong>and</strong> noisy <strong>and</strong> need regular<br />
maintenance. On the other h<strong>and</strong> steam savings of 60% are claimed <strong>and</strong>, in contrast to the<br />
TVR system (see below), the only hot water produced is from the heater condensate heat<br />
exchanger/cooler.<br />
The thermal vapour recompression systems (TVR) have the same objectives as the<br />
MVR systems, but vapour recompression is achieved using asteam jet compressor, an<br />
important consequence of which is that substantial amounts of hot water are generated.<br />
The economics of this system largely depend on whether this hot water is needed. In the<br />
compressor ajet of live steam, regulated by aneedle valve, enters achamber where it<br />
draws in vapour from the copper stack <strong>and</strong> mixes with it <strong>and</strong> carries it forward to an<br />
expansion chamber where it slows down <strong>and</strong> the pressure rises (Fig. 10.15). Values vary,<br />
but the live steam must have an adequate pressure, at least 8bar up to 24 bar. The<br />
compressed vapour has an overpressure of 0.3±0.4 bar <strong>and</strong> atemperature of 106±110ëC<br />
(222.8±230ëF). Agood proportion, preferably the major part, of the vapour (55±74%) is<br />
compressed<strong>and</strong>used tosupporttheboilinthecopperwhilethe remainderisdiverted toa<br />
condenser which heats cooling water to 80±95ëC (176±203ëF; Fig. 10.16).<br />
The vapour condensate joins the condensate from the wort heater <strong>and</strong> passes through a<br />
cooler, which generates warm water, <strong>and</strong> cool condensate, which may be used or be<br />
diverted to waste (Dymond <strong>and</strong> Djurslev, 1994; Fohr <strong>and</strong> Meyer-Pittroff, 1998). The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Valve<br />
actuator<br />
Vapours<br />
Live steam (providing<br />
motive power)<br />
Steam jet with<br />
regulator<br />
Steam jet<br />
Mixed vapours <strong>and</strong> steam<br />
pressure chamber<br />
Fig. 10.15 The principle of a steam-jet vapour compressor. A regulated jet of live steam sucks<br />
vapours into the nozzle. As the cross-sectional area of the tubing widens the stream of mixed steam<br />
<strong>and</strong> vapour slows down <strong>and</strong> the pressure rises, e.g., to 1.2±1.7 bar (104.8±115.2 ëC; approx.<br />
221±239 ëF).<br />
Heated<br />
water,<br />
80–95°C<br />
Vapour<br />
condenser<br />
Vapour<br />
condensate<br />
(hot)<br />
Cool<br />
water<br />
Cool water<br />
Heated water<br />
25–46%<br />
vapour<br />
Wort<br />
return<br />
Vapour<br />
Copper, 100°C,<br />
0.02 bar<br />
Throttle<br />
55–74%<br />
vapour<br />
Wort<br />
circulation<br />
pump<br />
Steam jet<br />
vapour<br />
compressor<br />
101–106°C<br />
External<br />
105–115°C<br />
cal<strong>and</strong>ria<br />
heater<br />
Condensate<br />
Mixed condensate (hot)<br />
Vapour<br />
condensate<br />
cooler<br />
Cooled, mixed<br />
condensate,<br />
20–30°C<br />
Effluent<br />
Motive steam,<br />
7–24 bar<br />
Mixed vapours<br />
<strong>and</strong> steam,<br />
1.2–1.7 bar<br />
Bypass<br />
Valve<br />
Fig. 10.16 A thermal vapour recompression (TVR) system (after Fohr <strong>and</strong> Meyer-Pittroff, 1998).<br />
The wort is boiled, under pressure, by the compressed mixture of vapour <strong>and</strong> steam from the steam<br />
jet vapour compressor. Vapour from the copper that is not compressed is condensed <strong>and</strong> used to<br />
heat water in the process. Mixed condensates from the vapour condenser <strong>and</strong> the heater are cooled<br />
<strong>and</strong> sent to waste. Hot water is also obtained from this operation.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
warm water from the condensate cooler may be used as the cooling water in the vapour<br />
condenser. Such systems use the vapour with 40±50% efficiency <strong>and</strong> produce hot water<br />
in amounts of 0.2±1.5 hl/hl of beer produced. TVR systems may be suitable for smaller<br />
<strong>and</strong> medium sized breweries. The investment in the steam jet compressor is relatively<br />
small, little maintenance is required <strong>and</strong> the compressor should run quietly. The system<br />
must be purged <strong>and</strong> air-free before it can operate. It must be free from leaks.<br />
10.8 Hot wort clarification<br />
At the end of the boil the wort should be absolutely clear (`bright') but contain,<br />
suspended in it, the remains of hops <strong>and</strong> flocs of trub or hot break. If whole hops are<br />
employed then the spent hops will probably weigh 0.7±1.4 kg/hl (2.4±5.0 lb./imp. brl) wet<br />
weight <strong>and</strong> will be associated with a significant amount of wort. The trub will be in the<br />
region of 0.21±0.28 kg/hl (0.75±1.0 lb./imp. brl) wet weight <strong>and</strong> will contain 80±85%<br />
water (Hough et al., 1982). Hot break contains roughly 50±60% crude protein, 20±30%<br />
tannin, 15±20% resins <strong>and</strong> 2±3% ash (dry wt., Andrews, 1992). Significant quantities of<br />
lipids are also present. Flocs of trub may reach 5±10 mm in diameter, but these can easily<br />
be disrupted, e.g., by pumping, into particles of 20±80 m diameter <strong>and</strong> a greater<br />
exposure to shear will reduce these to particles of 0.5±1.5 m. The hot break should be<br />
removed from the wort as thoroughly as possible, <strong>and</strong> this is most easily achieved with<br />
large particles. Consequently boiled wort should be h<strong>and</strong>led gently <strong>and</strong> shear should be<br />
avoided to minimize damage to the trub.<br />
The methods used to separate spent whole hop cones <strong>and</strong> the remains of milled or<br />
milled <strong>and</strong> pelleted hops are different. In older, small breweries the remains of the hops<br />
were removed from the wort, when it was cast from the copper, by straining it through a<br />
cloth bag or a sieve. In larger breweries the hops were sieved out of the wort using a<br />
Montejus or hop jack (Fig. 10.17). Some trub was retained with the spent hops, which<br />
Wort discharge<br />
Sparge ring<br />
Wort in<br />
Mesh screen<br />
Agitator<br />
Hops discharge<br />
Fig. 10.17 A Montejus, or hop jack, used for sieving whole hops from wort when it is discharged<br />
from the copper (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Spray ball<br />
for in-place<br />
cleaning<br />
Vent stack<br />
Spray plate<br />
Sparge arms<br />
Insulation<br />
Hot water for sparge<br />
Spray ball<br />
Wort main<br />
Water main<br />
Floor<br />
Slotted plates<br />
Underletting<br />
Spent hops<br />
discharge<br />
pipe Pump<br />
Wort<br />
Wort to<br />
cooling section<br />
recirculation<br />
pipe<br />
Fig. 10.18 Atraditional British hop back, for separating whole hops from cast wort (Hough et al.,<br />
1982).<br />
were sparged with hot water from fixed sparge-rings, to recover entrained extract. The<br />
remaining trub settled in the coolships, when these were in use. The spent hops were<br />
discharged mechanically. Amore refined device, used for separating hops, is the hop<br />
back (Fig. 10.18). In these the wortcontaining the remains ofwhole hops is spread over a<br />
false bottom in avessel that, in later versions, is enclosed. The wort is withdrawn from<br />
below the plates, having slots of about 1.55mm (0.06in.) occupying 25±30% of the base<br />
<strong>and</strong>,atfirst,isreturnedtothetopofthebedofhopsthatisformed.Thisfiltersoffmostof<br />
the hot break. When the wort is clear (`running bright') it is directed to the cooler. The<br />
process is comparatively slow. The vessel, which in modern designs is equipped for CIP,<br />
superficially resembles amashtun, <strong>and</strong>indeedin some small breweriesthe mashtuns are<br />
used as hop backs.<br />
To work well the bed of spent hops should be 30±60cm (c. 1±2ft.) deep; 15cm (6in.)<br />
is regarded as the minimum depth. As hopping rates have been reduced <strong>and</strong> the use of<br />
milled <strong>and</strong> pelleted hops has become commonplace, hop backs have become less<br />
common. Sometimes aroma hops are added to the hop back, to impart more `hoppy<br />
character' to abeer. From the hop back, the clarified wort istransferred, either directly or<br />
after brief storage in abuffer tank, to the cooling system. To recover the wort retained in<br />
the spent hops these are sparged with hot water (typically 8l/kg spent hops; about<br />
0.8gal./lb.). Finally, the drained hops are removed. Spent hops may be disposed of with<br />
the spent grains. Cattle accept the grains/hops mixture. Cleaning may take place after<br />
every brew or after several brews.<br />
Hop separators, (Fig. 10.19), drain the hops over astrainer, carry them forward <strong>and</strong><br />
then sparge them <strong>and</strong> compress them with either ascrew or abelt conveyor press,<br />
squeezing out residual liquid, before the spent hops are discharged. Separators do not<br />
separate most of the trub from the wort, <strong>and</strong> so trub removal has to be achieved in a<br />
separate operation.<br />
Increasingly hop powders, pelleted powders <strong>and</strong> extracts are used, <strong>and</strong> with these hop<br />
backs <strong>and</strong> separators cannot be employed. In these cases the wort may be clarified by hot<br />
filtration, by centrifugation, by sedimentation in acoolship or in amore modern settling<br />
tank. But the most commonly used device in newer breweries is the whirlpool tank or the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1<br />
3<br />
2<br />
2<br />
2<br />
12<br />
Fig. 10.19 Ahop separator (Hough et al., 1982, courtesy of A. Ziemann, GmbH). 1. Valve;. 2.<br />
Level control electrodes; 3. Level controller; 4. Helical screw conveyor; 5. Sieve/strainer; 6.<br />
Compression chamber, where the hops are squeezed. 7. Fasteners; 8. Wort receiver under the<br />
second strainer; 9. Three-way valve (wort return <strong>and</strong> sparge water); 10. Return pipe; 11. Wort<br />
discharge; 12. Clean-out valve. Spent hops are discharged at the top.<br />
combined copper (or kettle)/whirlpool.Filtration of the hot wort may be carried out using<br />
vertical, horizontal or c<strong>and</strong>le filters employing kieselguhr or Perlite filter aids. Filtration<br />
gives exceptionally clear worts, but wort losses can be significant, it is difficult to<br />
maintain sterility <strong>and</strong> there are the costs of purchasing the filter aids <strong>and</strong> disposing of the<br />
used aid <strong>and</strong> sludge. Coolships are now rarely or never used although their shallow liquid<br />
layers allowedeffectivesedimentation<strong>and</strong> wort clarification. Wortlosses<strong>and</strong>the risksof<br />
microbiological contamination were high.<br />
Thelogical successorstocoolshipsare sedimentationtanks(Fig. 10.20; Haecht, vanet<br />
al., 1990; Rehberger <strong>and</strong> Luther, 1994; Kunze, 1996; Vermeylen, 1962; Versteegh,<br />
1989). These may have flat or conical bases <strong>and</strong> the sides of the tanks, which are<br />
generallycylindrical, may ormay not be cooled by water inwall jackets. If cooling water<br />
is used it is itself heated <strong>and</strong> is retained for use. The vessels are enclosed <strong>and</strong> have<br />
vapour-escape stacks. Usually each vessel is filled from above (although in modern<br />
vessels bottom filling, to minimize oxygen pick-up <strong>and</strong> wort oxidation, would be<br />
expected) <strong>and</strong> is allowed to st<strong>and</strong> for about one hour, while solids in suspension<br />
progressively sediment. Wort is drawn from the top of the liquid, which clarifies first,<br />
using ahinged tube supported by afloat. The turbidity of the wort that is collected is<br />
continuously monitored <strong>and</strong> if the value rises too much collection is slowed or stopped.<br />
The trub <strong>and</strong> hop fragments form asloppy, `turbid wort' that is collected <strong>and</strong> processed<br />
separately. There has been aproposal to cool the edge of the surface of the resting wort<br />
by evaporation, to encourage the downward flow of the wort by the vessel wall, across<br />
thebaseofthevessel,whereprecipitatedmaterialsaccumulate,<strong>and</strong>upwardsinthecentre<br />
(Versteegh, 1989). The result claimed is aconical deposit of material in the middle of the<br />
flat base of the vessel analogous to that obtained in whirlpool vessels (see below).<br />
Sometimes the hot wort is clarified by centrifugation in adecanter or awort-clarifying<br />
centrifuge. These devices may also be used for wort recovery from wet trub samples. It is<br />
desirable, particularly if the wort is heavily hopped, to pre-treat it by passage through a<br />
rotary brush strainer <strong>and</strong> ahydrocyclone, to remove some of the hop solids <strong>and</strong> abrasive<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
11<br />
10<br />
4<br />
9<br />
5<br />
8<br />
6<br />
7
Air vent<br />
Lower wort<br />
collection tube<br />
Trub<br />
pump<br />
Valves<br />
Float<br />
Wort inlet<br />
Valve<br />
Trub outlet<br />
Valve<br />
Valve<br />
Hinged wort<br />
collection tube,<br />
with float<br />
Wort<br />
pump<br />
Fig. 10.20 An empty hot wort settling tank (after Rehberger <strong>and</strong> Luther, 1994). After aperiod of<br />
settling clear wort is drawn from the top, through the floating, hinged arm. At the end of the settling<br />
period the last of the clear wort is collected from the lower collection point <strong>and</strong> then the trub, mixed<br />
with wort, is also drawn off.<br />
materials, before it enters the centrifuge. Modern centrifuges discharge collected solids<br />
automatically, they give good wort recoveries (losses of
(a)<br />
Taylor eddy<br />
(b)<br />
Rotation<br />
Centrifugal<br />
force<br />
Trub cone<br />
Planetary<br />
eddy<br />
Wort outlet<br />
Surface of wort<br />
Desired pattern of<br />
wort circulation<br />
Torus eddy<br />
Level for inserting<br />
lattice or rings<br />
Fig. 10.21 Currents in whirlpool tanks. (a) The ideal flow pattern in a whirlpool. (b) Undesirable<br />
eddy currents that can occur <strong>and</strong> which interfere with trub separation. The most troublesome are the<br />
torus eddies, which may be checked by the insertion of a lattice or rings at the level indicated (after<br />
Denk, 1991).<br />
In the centre of the tank the liquid rises, leaving solids deposited. The vessel is usually<br />
filled in 10±15 minutes <strong>and</strong> after 20±30 minutes the wort has cleared <strong>and</strong> the solids are<br />
deposited as a cone in the centre of the base of the vessel, surrounded by a clear, annular<br />
space. Wort is typically withdrawn from three exit points, often about half-way down the<br />
vessel, three-quarters of the way down <strong>and</strong> in the base of the vessel towards the side,<br />
where the clear zone should be. This allows the wort to be drawn off at successively<br />
lower levels as it clears. Its turbidity is generally monitored automatically. One or more<br />
of the upper take-off points may withdraw wort tangentially to support the rotary motion.<br />
Sometimes new whirlpools do not work in a satisfactory way <strong>and</strong> must be modified.<br />
Empirical <strong>and</strong> theoretical studies have improved the situation. Problems arise quite<br />
separately from the design of the whirlpool <strong>and</strong>/or how it is operated <strong>and</strong> from problems<br />
with the trub caused by factors in the brewing process, which operate before the wort<br />
enters the whirlpool. In attempts to improve whirlpool performance a large number of<br />
types of vessel bases have been tried, including flat <strong>and</strong> level or slightly inclined (with or<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
without central `cups', or sumps of various sizes or a circumferential gully), inverted<br />
cones <strong>and</strong> conical bases (Andrews, 1988; Andrews <strong>and</strong> Axcell (private communication);<br />
Denk, 1991, 1994, 1998; Wilkinson, 1991 a, b).<br />
One problem is that secondary eddies in the wort can disrupt the desired flow pattern<br />
<strong>and</strong> impede the sedimentation of the trub <strong>and</strong> hop residues (Fig. 10.21b). Two solutions<br />
have been proposed; first the installation of an annular, horizontal metal grid or, secondly,<br />
a set of 4±5 circular metal rings 25±60 cm above the base of the vessel. In each case these<br />
devices impede the flows of the secondary torus eddies <strong>and</strong> greatly improve the deposition<br />
of the trub. These devices are installed only if experience shows that they are needed.<br />
Sumps interfere with the desired flow pattern <strong>and</strong> encourage the collection of as much as<br />
2% of the wort volume as trub wort, which must be processed separately to recover the<br />
entrained extract. The most popular configurations are now vessels with either flat bases,<br />
slightly inclined to the horizontal, to encourage drainage of the wort to the exit point at the<br />
outer edge of the clear zone, which surrounds the trub cone, or shallow, conical bases, with<br />
angles of about 30ë (Fig. 10.22). In the first case, the cone remaining after the vessel has<br />
been drained should be dry. Subsequently it must be dislodged <strong>and</strong> dispersed, for example,<br />
with pulses of water from rotating, electrically driven jets, so that it drains to the lowest<br />
wort outlet. This cleaning should not be long delayed as the exposed trub cone can set hard<br />
in air. In the other case the trub, still with some wort, is withdrawn from the apex of the<br />
cone, <strong>and</strong> the wort is recovered by centrifugation.<br />
Problems with the trub cone may be caused at earlier stages in the brewing process by<br />
using an abnormal grist, by having inadequate trub flocculation in the boil or by breaking<br />
up the trub through shearing the wort during pumping, by using a high-gravity wort<br />
which, because of its high density `buoys-up' the trub, <strong>and</strong> so on. Attempts to improve<br />
cone formation have included the exclusion of air, additions of copper finings, nylon or<br />
PVPP powders, kieselguhr, bentonite <strong>and</strong> KMS (potassium metabisulphite). As the wort<br />
is withdrawn the trub cone is exposed <strong>and</strong> so is no longer supported by the buoyancy<br />
provided by surrounding wort. It then tends to slump <strong>and</strong> spread outwards <strong>and</strong> even break<br />
Working level<br />
Trub<br />
20°<br />
Vapour<br />
stack<br />
1:40 slope (1.5°)<br />
Trub jetting<br />
machine<br />
Tangential<br />
inlet<br />
Outlet 1,<br />
at 50%<br />
Outlet 2,<br />
at 10%<br />
Outlet 3,<br />
at trub<br />
Trub outlet<br />
(a) (b)<br />
20°<br />
Working level<br />
120°<br />
Vapour<br />
stack<br />
Outlet 3,<br />
at VT<br />
Tangential<br />
inlet<br />
Outlet 1,<br />
at 50%<br />
Outlet 2<br />
at 10%<br />
Fig. 10.22 Two patterns of kettle-whirlpools (Courtesy of Briggs of Burton, plc.). In each case<br />
boiling wort, driven by the thermosyphon effect generated in an external wort heater, is discharged<br />
tangentially above the surface of the wort. Volatiles readily escape with the vapour, <strong>and</strong> the wort<br />
rotates in the vessel driven by the incoming wort. (a) In this flat-based pattern direct wort recovery<br />
is favoured. The trub cone is drained <strong>and</strong> then, after wort recovery, is driven out with water jets. (b)<br />
In this pattern, with a conical base, trub is recovered as a slurry in wort, which is normally<br />
recovered. This design favours high wort clarity.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
,
up, so it may reach the lowest wort collection point. To counteract this, wort withdrawal<br />
is slowed when the trub cone is just exposed so that the trub is slowly drained <strong>and</strong> is<br />
exposed more slowly to the air, <strong>and</strong> the cone retains its shape better.<br />
Kettle-whirlpools are of various types. Some are equipped with internal heaters <strong>and</strong><br />
wortiswithdrawn <strong>and</strong>pumpedback viaaninjectionentrytoinducerotationintheliquid.<br />
This design requires the continuous use of apump, with the equipment <strong>and</strong> maintenance<br />
costs <strong>and</strong> disrupting effects on the trub that this entails. The internal heater clutters the<br />
centre of the vessel. Abetter design uses an external heater that circulates the boiling<br />
wort using the thermosyphon effect. The boiling wort is returned to the top of the vessel<br />
at or above the wort surface <strong>and</strong>, being delivered through atangential entry, drives the<br />
rotation of the liquid (Fig. 10.22). It can be operationally more convenient to have two<br />
copper-whirlpool vessels rather than a single copper <strong>and</strong> a single, separate whirlpool.<br />
The wet trub produced by hot wort clarification contains, before washing, about 80%<br />
of wort. If this is sent to drain its high BOD <strong>and</strong> large amount of suspended solids give<br />
rise to high effluent charges. In addition extract is wasted. Disposal with the spent grains<br />
also wastes extract. If the brewing schedule permits it, extract is most easily recovered by<br />
transferring the trub to the mashing vessel or lauter tun making the next brew. Sometimes<br />
clear wort has been recovered from trub by filtration or by centrifugation (Hansen et al.,<br />
1990). It is less costly to centrifuge the trub from a whirlpool than to omit the whirlpool<br />
<strong>and</strong> centrifuge all the wort. Cloudy wort has been obtained by vibrating screen filtration<br />
of trub, (Fig. 10.23), <strong>and</strong> this treatment has been used to pre-treat trub to provide the feed<br />
for complete clarification by cross-flow filtration (Maule et al., 1989; Visscher et al.,<br />
1991). The cross-flow filtration was carried out with ceramic filter units with notional<br />
0.2 m pores. Their use increased the brewhouse yield by 1%. However, the units were<br />
easily fouled, reducing the filtration rate <strong>and</strong> extract recovery. To minimize this problem<br />
the re-circulating trub/wort was diluted in the later stages of treatment.<br />
After the boil wort may remain in the whirlpool for up to an hour before being cooled.<br />
During this warm period chemical changes continue. Some, like the isomerization of the<br />
Mesh screen<br />
Clarified<br />
material<br />
out<br />
Eccentric<br />
weights<br />
Feed material<br />
in<br />
Solid<br />
material<br />
out<br />
Motor<br />
Solids<br />
out<br />
(a)<br />
(b)<br />
Fig. 10.23 (a) A vertical section of a vibrating screen filter <strong>and</strong> (b) the distribution <strong>and</strong> movement<br />
of solids on the filter surface as seen from above (Button et al., 1977).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
-acids, are desirable but others, such as the generation of some volatile Maillard<br />
reaction products <strong>and</strong> DMS, may not be. The proposal to slow these changes by cooling<br />
the wort alittle as it enters the whirlpool has already been noted. Another approach,<br />
already noted, is to `strip' the wort of volatiles as it leaves the whirlpool, just before it is<br />
delivered to the cooling system. This has been achieved in different ways, for example,<br />
by spraying the wort into avacuum chamber, by thin-film evaporation or by steamstripping.<br />
Similar effects occur in the expansion/cooling chambers in high-temperature,<br />
continuous wort boiling (Section 10.5).<br />
10.9 Wort cooling<br />
After clarification the hot wort must be cooled to the temperature at which it is pitched<br />
(inoculated) with yeast. Traditionally this is about 15±22ëC (59±71.6ëF) for ales <strong>and</strong> 6±<br />
12ëC (42.8±53.6ëF) for lagers, but other temperatures are used. The cooling should be<br />
carriedoutrapidly<strong>and</strong>underasepticconditionstostopchemicalreactionscontinuing<strong>and</strong><br />
to minimize chances of growth of any contaminating microbes. As the wort cools it<br />
becomeshazy asacoldbreakforms. Thismayormaynotberemoved(Section10.10).In<br />
addition the wort must be charged with oxygen to an appropriate level (Section 10.11;<br />
Chapters 11 <strong>and</strong> 12). In modern breweries the heat from the hot wort is partly recovered<br />
in hot water.<br />
The oldest coolers were the open, horizontal coolships, (Section 10.8), in which the<br />
wort was held in rectangular hardwood, copper, iron or aluminium trays in ashallow<br />
layer of, say, 15±25cm (c. 6±10in.), for as long as 12 hours, to cool, to pick up oxygen<br />
from the air <strong>and</strong> to deposit suspended materials, including hop fragments, hot break <strong>and</strong><br />
some cold break. Ideally, the room holding the coolship had acurved ceiling designed to<br />
preventcondensation drippingintothewort<strong>and</strong>,inlatertimes,wasventilated withsterile<br />
air (Hind, 1940; Sykes <strong>and</strong> Ling, 1907; Vermeylen, 1962). In <strong>practice</strong> coolships were<br />
difficult to keep clean <strong>and</strong> free from microbiological contamination, they took up much<br />
space <strong>and</strong> wort losses were high. In some cases cooling was accelerated by passing cold<br />
water through pipes placed on the base of the coolship, an arrangement that accelerated<br />
cooling, but with the added disadvantage of making the equipment awkward to clean.<br />
The introduction of vertical coolers was a substantial advance, since they were<br />
compact, easier to clean, <strong>and</strong> cooled the wort comparatively rapidly. The wort was<br />
introduced to atrough at the top, from which it trickled down as athin film over aseries<br />
ofhorizontal tubeswithin which cold waterflowedupwards, inacounter-flow. Thetubes<br />
mightbeofvariouscross-sectionsorhaveprojectionsonthemdesignedtodirecttheflow<br />
of the wort <strong>and</strong> to increase the area of the heat-transfer surfaces. At the base of the cooler<br />
the wort was collected in atrough <strong>and</strong> piped to afermenter. The wort was aerated in this<br />
process, but the cold break was not removed.<br />
Sometimeswortiscooled,inclosedshell<strong>and</strong>tubecoolers,bypassingitdownthrough<br />
stainless steel tubes, which are successively cooled by air in the upper part, water in the<br />
central part <strong>and</strong> refrigerant in the lower part. Sterile air is injected into the base <strong>and</strong> rises<br />
upthetubesinacounterflowtothewort.Aunitabletocool450hl/h(275brl/h)wouldbe<br />
about 7m(23ft.) tall <strong>and</strong> 2m(6ft. 6in.) wide (Hough et al., 1982). By far the most<br />
common coolers are plate heat exchangers (paraflows), which are compact, efficient <strong>and</strong><br />
versatile. Being enclosed microbes are excluded, <strong>and</strong> cleaning <strong>and</strong> sterilization are<br />
simple. In these coolers numerous stainless steel plates are suspended vertically from a<br />
strong metal frame <strong>and</strong> are clamped together (Fig. 10.24). The numbers of plates can be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
Hot wort<br />
in (93°C)<br />
(b)<br />
Water out (68°C) Water in (16°C) Glycol out Glycol in<br />
(–1°C)<br />
Cool wort<br />
out (10°C)<br />
Frame supporting plates<br />
<strong>and</strong> holding them together<br />
Fig. 10.24 (a) The principles of a plate <strong>and</strong> frame heat exchanger, indicating the movement of the<br />
liquid to be cooled <strong>and</strong> the counterflow of the cooling refrigerant (Hough et al., 1982). (b) The<br />
flows in a two-stage cooling system (after Hough, 1985). In the first stage wort is cooled by a<br />
counterflow of water, which is heated in the process. In the second stage the wort is cooled further<br />
by a chilled glycol refrigerant.<br />
varied <strong>and</strong>, as a result, so can the exchangers' cooling capacity. The plates are indented<br />
with complex patterns <strong>and</strong> have four holes in the corners. The edges of the plates <strong>and</strong> the<br />
holes are rimmed with rubber gaskets so that when the plates are pressed together the<br />
gaskets seal, <strong>and</strong> a series of ducts <strong>and</strong> channels are formed. In operation the wort flows<br />
through one channel, across the stack, between two plates <strong>and</strong> out to another channel<br />
while cold water or refrigerant flows in the opposite direction (`counter-current') in<br />
alternating gaps between the plates. The indented patterns on the plates ensure turbulent<br />
flow <strong>and</strong> efficient heat exchange.<br />
The coolants can be supplied in various ways. For example, water should be the first<br />
so that it can be heated, to 70 ëC (158 ëF) or more, <strong>and</strong> used around the brewery. Part way<br />
through the cooler, cooling may be supplied by chilled water or a glycol, ammonia or an<br />
alcohol refrigerant. The arrangement used will be decided, in part, by the final wort<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
temperature required. However, if acoolant other than water is used care is needed to<br />
regularly check the integrity of the plates <strong>and</strong> seals, so that there is no possibility that a<br />
coolant can leak <strong>and</strong> contaminate the wort. Coolers must be regularly cleaned, to remove<br />
scale<strong>and</strong> fouling, <strong>and</strong> must be checked for leaks. The plates are mounted toallow ease of<br />
separation <strong>and</strong> replacement. Sometimes sterile air or oxygen is injected into the wort in<br />
the cooler to take advantage of the turbulent flow that encourages the gas to dissolve.<br />
10.10 The cold break<br />
As the wort cools it becomes cloudy as the cold break or trub separates from solution.<br />
This material contains about 50% protein, 15±25% polyphenols <strong>and</strong> 20±30% of wort<br />
carbohydrates (see Chapter 9). Unlike the hot break this material does not flocculate, <strong>and</strong><br />
occurs as small particles,
Cold break (mg/ml)<br />
300<br />
250<br />
200<br />
150<br />
100<br />
50<br />
0<br />
10.11 Wort aeration/oxygenation<br />
50 68 86 104 122 140 158 176 194<br />
Temperature (°F)<br />
0 10 20 30 40 50 60 70 80 90<br />
Temperature (°C)<br />
Fig. 10.25 The amounts of cold trub (break) formed in aparticular wort at various temperatures<br />
(Hough et al., 1982).<br />
In the initial stage of `fermentation' the freshly pitched yeast needs to be in wort that<br />
containsdissolvedoxygen.Theconcentrationofoxygenrequirediscritical,<strong>and</strong>dependson<br />
the wort, for example, the availability of sterols <strong>and</strong> unsaturated fatty acids, <strong>and</strong> the variety<br />
<strong>and</strong> history of the yeast (Chapters 11 <strong>and</strong> 12). In the past it seems that saturating wort with<br />
oxygen from air (approx. 21% O 2)was sufficient, but now saturation with pure oxygen is<br />
often required. It is surprisingly difficult to dissolve oxygen quickly in aqueous solutions.<br />
At equilibrium the amount of gas dissolved at achosen temperature is proportional to the<br />
partial pressure of the gas above the liquid. Although solubilities are often reported at a<br />
st<strong>and</strong>ard atmospheric pressure, (equal to that at the base of acolumn of mercury 760mm<br />
(29.92in.) high at 0ëC (32ëF), at sea level at 45ë latitude), decreasing or increasing this<br />
pressure will proportionally change the equilibrium concentration of the gas in solution.<br />
Atmospheric pressure is always fluctuating. As the temperature increases, so the amount of<br />
oxygen in solution, in equilibrium with air or pure oxygen, declines (Table 10.4).<br />
Dissolving substances (salts, sugars, etc.) in water reduces the amount of gas that can<br />
be dissolved. Wort is astrong solution of amixture of substances <strong>and</strong> so the solubility of<br />
oxygen in wort is less than in pure water (Table 10.4), <strong>and</strong> the stronger, more<br />
concentrated the wort the less oxygen will dissolve in it under afixed set of conditions.<br />
Wort may be injected with air or oxygen at the inlet to the cooler, part-way through the<br />
cooler (Section 10.9), or after the cooling process is complete. Adding the gas to hot or<br />
warm wort allows the oxidation of wort components, causing flavour changes <strong>and</strong><br />
darkening, which are usually undesirable. However, adding the gas to the hot wort means<br />
that the sterility of the gas is not so critical as microbes will be killed at the elevated<br />
temperatures. Air or oxygen added to cooled wort must have been sterilized, either by<br />
filtration, e.g., through sintered metal with pores
Table 10.4 The solubility of oxygen (mg O2/l) from the air in water or wort at different<br />
temperatures <strong>and</strong> from pure oxygen gas, all at a st<strong>and</strong>ard atmospheric pressure (data of Krauss,<br />
1967). Compare with Appendix A.12.<br />
Temperature From air From oxygen<br />
(ëC) (ëF) Water Wort (12%) Water<br />
0 32 14.5 11.6 69<br />
3 37.4 ± ± 64<br />
5 41 12.7 10.4 61<br />
8 46.4 ± ± 56<br />
10 50 11.2 9.3 54<br />
15 59 10.0 8.3 48<br />
20 68 9.9 7.4 ±<br />
The rate of dissolution of a gas in liquid depends on the pressure, the thoroughness of<br />
mixing <strong>and</strong> the area of the gas/liquid interface. Consequently, it is most efficient to direct<br />
a stream of very fine gas bubbles, under pressure, into a turbulent flow of wort. Devices<br />
used for aeration/oxygenation include sintered metal or ceramic c<strong>and</strong>les, which release<br />
clouds of very fine bubbles into the base of a vessel or into a flowing stream of wort <strong>and</strong><br />
are strong but which are difficult to keep clean, <strong>and</strong> centrifugal mixers which are very<br />
efficient but are expensive (Kunze, 1996). Devices based on the Venturi tube principle,<br />
which aerate/oxygenate cooled wort while flowing to a fermenter, are common. The<br />
flowing wort comes to a restriction in the pipework, which causes it to accelerate <strong>and</strong> the<br />
pressure to drop. At this point fine bubbles of air or oxygen are introduced into the liquid<br />
either from a fine nozzle, discharging into the stream of wort, or from fine perforations or<br />
sintered material in the tube wall. The clouds of bubbles are carried forward into the next<br />
section where the pipe abruptly exp<strong>and</strong>s, the flow slows <strong>and</strong> so the pressure rises <strong>and</strong> the<br />
flow becomes turbulent, conditions which favour the rapid solution of the gas. Sometimes<br />
the pipework downstream from the gas injection point contains inserts of metal `tapes' or<br />
net-like units that act as mixers, creating turbulence in the wort flow.<br />
In modern plant the dissolution of added oxygen is nearly complete. Oxygenation is<br />
automatically controlled, the rate of supply of gas is continuously monitored, for<br />
example, by measuring the flow rate or the decline in weight of gas cylinders.<br />
Alternatively, mass flow meters may be used. Sometimes oxygen is added to the first part<br />
of the wort flow <strong>and</strong> yeast is added to the second part. The level of dissolved oxygen,<br />
(really the equilibrium partial pressure), is usually monitored continuously by a<br />
membrane-covered oxygen electrode or cell.<br />
10.12 References<br />
ANDREWS, J. M. H. (1988) Ferment, 1 (3), 47.<br />
ANDREWS, J. M. H. (1992) Proc. 22nd Conv. Inst. of <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Section),<br />
Melbourne, p. 65.<br />
ANDREWS, J. M. H. <strong>and</strong> AXCELL, B. C. Private communication.<br />
ANON. (1994) J. Inst. <strong>Brewing</strong>, 100, 130.<br />
BENITEZ, J. L., FORSTER, A., DE KEUKELEIRE, D., MOIR, M., SHARPE, F. R., VERHAGEN, L. C. <strong>and</strong> WESTWOOD,<br />
K. T. (1997) EBC Manual of Good Practice: Hops <strong>and</strong> Hop Products. 185 pp. NuÈrnberg, GetraÈnke-<br />
Fachverlag Hans Carl.<br />
BOER, F. P. DE (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; Wort boiling <strong>and</strong><br />
clarification. Strasbourg, p. 193.<br />
BONACCHELLI, B., HARMEGNIES, F. <strong>and</strong> GIL, R. T. (2001) Proc. 28th Congr. Eur. Brew. Conv., Budapest<br />
(CD, paper 24; p. 235).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
BOYES, R. (1993) Brewers' Guard., 122 (3), 23.<br />
BRAEKELEIRS, R. <strong>and</strong> BAUDUIN, C. L. (2001) Proc. 8th Brew. Conv., Inst. of <strong>Brewing</strong> (Central <strong>and</strong><br />
Southern African Section), Sun City, p. 36.<br />
BUTTON, A. H., STACEY, A. J. <strong>and</strong> TAYLOR, B. (1977) Proc. 16th Congr. Eur. Brew. Conv., Amsterdam, p. 377.<br />
CHANTRELL, N. S. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, p. 89.<br />
CHANTRELL, N. S. (1984) MBAA Tech. Quart., 21 (4), 166.<br />
CLARKE, M. H. M. <strong>and</strong> KERR, R. A. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium: Wort<br />
boiling <strong>and</strong> clarification, Strasbourg, p. 139.<br />
COUTTS, M. W., RICKETTS, J. <strong>and</strong> SELKIRK, R. C. (1955) Proc. Conv. Master Brewer's Assoc. Amer., p. 20.<br />
CROMPTON, I. E. <strong>and</strong> HEGARTY, P. K. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, p. 625.<br />
DENK, V. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium: Wort boiling <strong>and</strong> clarification.<br />
Strasbourg, p. 155.<br />
DENK, V. (1994) Ferment, 7 (5), 299.<br />
DENK, V. (1998) Brauwelt Internat., 16 (1), 31.<br />
DICKEL, T., KROTTENTHALER, M. <strong>and</strong> BACK, W. (2002) Brauwelt Internat., 20 (1), 23.<br />
DYMOND, G. <strong>and</strong> DJURSLEV, O. (1994) Brew. Distill. Internat., 25 (4), 16.<br />
FOHR, M. <strong>and</strong> MEYER-PITTROFF, R. (1998) Brauwelt Internat., 16 (4), 304.<br />
FORREST, I. S., SKRGATIC, D., COKER, I. A. J. <strong>and</strong> HEAP, J. (1993) Proc. 24th Congr. Eur. Brew. Conv., Oslo,<br />
p. 493.<br />
FORTUIN, B. (1995) The Brewer, 81, 443.<br />
GRASMAN, R. <strong>and</strong> VAN EERDE, P. (1986) Proc. 19th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong><br />
Section), Hobart, p. 161.<br />
HACKENSELLNER, T. (1999) Brauwelt Internat., 17 (6), 495.<br />
HAECHT, J.-L., VAN, DE BRACKELEIRE, C., DUFOUR, J.-P. <strong>and</strong> DEVREUX, A. (1990) Eur. Brew. Conv.<br />
Monograph-XVI. EBC Symposium, `Separation Processes', Leuven, p. 96.<br />
HALL, R. D. <strong>and</strong> FRICKER, R. (1966) Proc. 9th Conv. Inst. <strong>Brewing</strong> (Australia <strong>and</strong> New Zeal<strong>and</strong> Section),<br />
Auckl<strong>and</strong>, p. 45.<br />
HANCOCK, J. C. <strong>and</strong> ANDREWS, J. M. H. (1996) Ferment, 9 (6), 344.<br />
HANSEN, N. L., LUND, M. <strong>and</strong> OLSEN, N. O. (1990) Eur. Brew. Conv. Monograph-XVI. EBC Symposium<br />
`Separation Processes', Leuven, p. 84.<br />
HERRMANN, H. (1985) Brew. Distill. Internat., Mar., p. 32.<br />
HERRMANN, H. (1998a) The Brewer, 84 (1005), 333.<br />
HERRMANN, H. (1998b) Ferment, 11 (1), 36.<br />
HERRMANN, H. (1999) Ferment, 12, Feb/Mar., 36.<br />
HIND, H., L. (1940) <strong>Brewing</strong> Science <strong>and</strong> Practice, II. <strong>Brewing</strong> Processes. London, Chapman <strong>and</strong> Hall,<br />
pp. 507±1020.<br />
HOUGH, J. S. (1985) The Biotechnology of Malting <strong>and</strong> <strong>Brewing</strong>. Cambridge University Press. 169 pp.<br />
HOUGH, J. S., BRIGGS, D. E., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1982) Malting <strong>and</strong> <strong>Brewing</strong> Science. II.<br />
Hopped Wort <strong>and</strong> Beer. London, Chapman <strong>and</strong> Hall, pp. 389±914.<br />
HUDSTON, H. (1969) MBAA Tech. Quart., 6 (1), 164.<br />
KANTELBERG, B., WIESNER, R., JOHN. L. <strong>and</strong> BREITSCHOPF, J. (2000) Brew. Distill. Internat., 31 (3), 16.<br />
KOLLNBERGER, P. (1986) MBAA Tech. Quart., 23, 126.<br />
KOLLNBERGER, P. (1987) Brauwelt, 127, 254.<br />
KRAUSS, G. (1967) Proc.11th Congr. Eur. Brew. Conv., Madrid, p. 35.<br />
KROTTENTHALER, M. <strong>and</strong> BACK, W. (2001) Brew. Distill. Internat., 32 (3), 14.<br />
KROTTENTHALER, M., HARTMANN, K. <strong>and</strong> BACK, W. (2001) Brauwelt Internat., 19 (6), 457.<br />
KUNZE, W. (1996) Technology <strong>Brewing</strong> <strong>and</strong> Malting. (Internat. edn, translated Wainwright, T.) Berlin,<br />
VLB, p. 254.<br />
LANGENHAN, R. (1995) J. Inst. <strong>Brewing</strong>, 101, 230.<br />
LENZ, B., LANGENHAN, R., HERRMANN, H., KANTELBERG, B. C. <strong>and</strong> FELGENTRAEGER, W. (1991) Proc. 3rd<br />
Sci. Tech. Conv. Inst. of <strong>Brewing</strong>, (Central <strong>and</strong> Southern African Section), Victoria Falls, p. 101.<br />
LUSTIG, S., KUNST, T. <strong>and</strong> HILL, P. (1997) Proc. 21st Congr. Eur. Brew. Conv., Maastricht, p. 341.<br />
MANGER, H.-J. (1998) Brauwelt Internat., 16 (4), 320.<br />
MAULE, D. R. <strong>and</strong> CLARK, B. E. (1985) Proc. 20th Congr. Eur. Brew. Conv., Helsinki, p. 379.<br />
MAULE, D. R., STEAD, J. R. <strong>and</strong> CLARK, B. E. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 393.<br />
MICHEL, R. A. (1991) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; `Wort boiling <strong>and</strong><br />
clarification'. Strasbourg, p. 118.<br />
MIEDANER, H. (1986) J. Inst. <strong>Brewing</strong>, 92, 330.<br />
MIEDANER, H. <strong>and</strong> NARZISS, L. (1986) Eur. Brew. Conv. Monograph-XI. EBC Symposium; `Wort<br />
Production'. Maffliers, p. 80.<br />
MITANI, Y., SUZUKI, H., ABE, T., NOMURA, M. <strong>and</strong> SHINOTSUKA, K. (1999) Proc. 27th Congr. Eur. Brew.<br />
Conv., Cannes, p. 619.<br />
MORRIS, D. R. (1987) Brew. Distill. Internat., 17 (3), 22.<br />
NARZISS, L. (1986a) Eur. Brew. Conv. Monograph-XI. EBC Symposium; `Wort Production'. Maffliers,<br />
p. 98.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
NARZISS, L. (1986b) Brauwelt, 126 (32), 1419.<br />
NARZISS, L. (1992) Die Bierbrauerei. II. Die Technologie der WuÈrzebereitung. (7th edn). Stuttgart,<br />
Ferdin<strong>and</strong> Enke. 402 pp.<br />
NARZISS, L. (1993) Proc. 4th Sci. Tech. Conv. Inst. of <strong>Brewing</strong>, (Central <strong>and</strong> Southern African Section),<br />
Somerset West, p. 195.<br />
NARZISS, L., KIENINGER, H. <strong>and</strong> REICHENDER, E. (1971) Proc. 13th Congr. Eur. Brew. Conv., Estoril,<br />
p. 197.<br />
NARZISS, L., MIEDANER, H. <strong>and</strong> SCHNEIDER, F. (1992) Brauwelt Internat., IV, 346.<br />
ORMROD, I. H. L. (1986) J. Inst. <strong>Brewing</strong>, 92, 131.<br />
REED, R. J. R. (1992) Ferment, 5 (2), 125.<br />
REHBERGER, A. J. <strong>and</strong> LUTHER, G. E. (1994) in H<strong>and</strong>book of <strong>Brewing</strong> (Hardwick, W. A. ed.), New York,<br />
Marcel Dekker, p. 247.<br />
ROYSTON, M. G. (1971) in Modern Brewery Technology. (Findley, W. P. K. ed.). London, Macmillan <strong>and</strong><br />
Co., p. 60.<br />
SCHU, G. E. (1995, Oct). Brauwelt Internat., 13 (4), 316.<br />
SCHWILL-MIEDANER, A. <strong>and</strong> MIEDANER, H. (2002) Brauwelt Internat., 20 (1), 19.<br />
SELDESLACHTS, D., VAN DER EYNDE, E. <strong>and</strong> DEGELIN, L. (1997) Proc. 26th Congr. Eur. Brew. Conv.,<br />
Maastricht, p. 323.<br />
STIPPLER, K. (2000) Ferment, 13 (1), 34.<br />
STIPPLER, K., WASMUHT, K. <strong>and</strong> GATTERMEYER, P. (1997) Brauwelt Internat., 15 (4), 358.<br />
SYKES, W. J. <strong>and</strong> LING, A. R. (1907) The Principles <strong>and</strong> Practice of <strong>Brewing</strong>. (3rd edn). London, Charles<br />
Griffin <strong>and</strong> Co., p. 496.<br />
UNTERSTEIN, K. (1992) Brauwelt Internat., 10 (1), 65.<br />
VERMEYLEN, J. (1962) Traite de la Fabrication du Malt et de la BieÁre. 2, pp. 861, 937. G<strong>and</strong>. Assoc.<br />
Royale des Anciens EleÁves de l'Institute SupeÂrieur des Fermentations.<br />
VERSTEEGH, C. W. (1989) Proc. 22nd Congr. Eur. Brew. Conv., Zurich, p. 291.<br />
VISSCHER, H. J., TETTELAAR, M. E. <strong>and</strong> MARTENS, F. B. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon,<br />
p. 649.<br />
VOLLHALS, B. (1994) MBAA Tech. Quart., 31 (1), 1.<br />
WILKINSON, N. R. (1985) Proc. 1st Sci. Tech. Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern African<br />
Section), Johannesburg, p. 188.<br />
WILKINSON, N. R. (1991a) Eur. Brew. Conv. Monograph-XVIII. EBC Symposium; `Wort Boiling <strong>and</strong><br />
Clarification'. Strasbourg, p. 100.<br />
WILKINSON, N. R. (1991b) Ferment, 4 (6), 388.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
11<br />
Yeast biology<br />
11.1 Historical note<br />
The abilities of Saccharomyces yeast to leaven dough in baking <strong>and</strong> generate ethanol for<br />
beverage production have been utilized by mankind for several millennia, albeit for most<br />
of that time, unwittingly. The consumption of alcoholic beverages is common to all<br />
civilizations. Production of the earliest drinks was probably serendipitous since natural<br />
sources of sugars are invariably contaminated with yeast. Metabolism of sugars by<br />
Saccharomyces yeasts results in the formation of ethanol <strong>and</strong> carbon dioxide even under<br />
aerobic conditions. The mind-altering effects of ethanolic beverages must have provided<br />
the primary motive for their continued consumption <strong>and</strong> a powerful impetus for empirical<br />
development of a controlled process of production. Coincidentally, such drinks would<br />
have been nutritious <strong>and</strong> a useful means of sanitizing potentially dangerous water supplies.<br />
<strong>Brewing</strong> of beer has its probable origins in the Middle East at some time between 6000<br />
<strong>and</strong> 8000 BC, where it apparently developed in t<strong>and</strong>em with organized agriculture<br />
(Corran, 1975). It seems that the use of cereals for baking <strong>and</strong> brewing developed<br />
simultaneously. Clearly, this must also have included the discovery of malting <strong>and</strong> the<br />
use of yeast for leavening of dough <strong>and</strong> fermentation. These activities became large-scale<br />
undertakings, for example, it has been reported that in ancient Mesopotamia 40% of<br />
cereals were cultivated purely for brewing (Corran, 1975).<br />
<strong>Brewing</strong> spread from the Middle East to become the dominant alcoholic beverage of<br />
northern Europe. Thus, Tacitus commented that beer was the common drink in Germany<br />
at the time of the Roman Empire (King, 1947). It was from old German that the word<br />
yeast arose, probably from gischen descriptive of the foaming or frothing during<br />
fermentation. Pliny described the use by the Gauls for leavening of dough of `foam'<br />
obtained from beer fermentations (King, 1947). It is implicit in this etymology that the<br />
importance of yeast to brewing <strong>and</strong> baking was appreciated in historical times although<br />
its vital nature was unrecognized. For example, in medieval Engl<strong>and</strong> the yeast crop<br />
resulting from fermentation was known as godisgood (Forget, 1988). Nevertheless, the<br />
earliest manifestation of the German Reinheitsgebot or beer purity laws introduced to<br />
Bavaria in 1493 by Duke Albrecht IV stipulated the exclusive use of hops, malted barley<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>and</strong> water for brewing. Yeast was not included as an ingredient since it was unknown<br />
(Narziss, 1984).<br />
The first description of individual yeast cells was published in 1680 by Antonie van<br />
Leeuwenhoek (Chapman, 1931). Using a primitive microscope he recorded the<br />
appearance of yeast flocs in fermenting wort <strong>and</strong> modelled the same in wax. He also<br />
noted the formation of gas bubbles but did not appreciate that this was a by-product of<br />
yeast metabolism. The realization of the biological nature of fermentation forms the<br />
basis of the development of modern biochemistry <strong>and</strong> microbiology. Early alchemists<br />
observed the vigorous gas formation that accompanied the metabolism of wort by yeast<br />
<strong>and</strong> coined the term fermentation from the Latin fevere, to boil. The alchemist view of<br />
fermentation was that this was a process of active separation. Thus, ethanol was present<br />
but could not be detected until the `impurities' yeast <strong>and</strong> carbon dioxide were removed<br />
(Florkin, 1972).<br />
Gay-Lussac in 1810 established the stoichiometry of ethanolic fermentation by<br />
demonstrating that two molecules each of ethanol <strong>and</strong> carbon dioxide devolved from each<br />
molecule of sugar consumed. He suggested that fermentation was initiated by exposure to<br />
oxygen since heated foodstuffs in sealed containers spoilt only when air was admitted. He<br />
also noted that spoilage could be stopped by further heating but failed to form the<br />
conclusion that there was a heat-labile component present. He suggested that heat<br />
changed the `ferment' to an inactive form in which it was impervious to the stimulating<br />
effects of oxygen. Yeast was not considered to play any part in fermentation on the basis<br />
of insolubility.<br />
Undoubtedly, early practitioners of brewing <strong>and</strong> baking appreciated the essential<br />
requirement for yeast in fermentation by dint of empirical observation. Later<br />
incarnations of the Reinheitsgebot included yeast as an ingredient in beer making<br />
based on the observations that addition to wort of some of the crop from a previous<br />
fermentation gave a more consistent process (Narziss, 1984). The role of yeast in<br />
fermentation <strong>and</strong> its vital nature was established independently by the Germans,<br />
Theodor Schwann (1837) <strong>and</strong> Friedriech Traugott KuÈtzing <strong>and</strong> a Frenchman, Charles<br />
Cagniard-Latour (1836). Both KuÈtzing <strong>and</strong> Cagniard-Latour used direct microscopic<br />
observation to describe yeast cells. The latter recorded yeast proliferation by the<br />
formation of buds <strong>and</strong> both deduced that the cells were living <strong>and</strong> had an active role in<br />
fermentation.<br />
Schwann's investigations were directed towards disproving the theory of spontaneous<br />
generation of life. He demonstrated that a meat infusion did not putrefy if heated.<br />
Furthermore, the onset of putrefaction subsequently brought about by the admission of air<br />
was prevented if the latter was first heated. He was able to demonstrate that heating of air<br />
did not change its essential character since it could still sustain the life of a frog. This<br />
disproved the earlier assertions of Gay-Lussac. Schwann also made microscopic<br />
observations of growing yeast cells, which he termed Zuckerpilz.<br />
Despite the observations of Schwann, the vital nature of yeast was not generally<br />
accepted until the work of Louis Pasteur was published (Anderson, 1995). Pasteur studied<br />
optically active molecules, which he considered were produced only by living organisms.<br />
He isolated optically active amyl alcohol from fermentations <strong>and</strong> therefore assumed that<br />
the process must be animate. Pasteur extended the work of Schwann <strong>and</strong> demonstrated<br />
that putrefaction of foodstuffs occurred via contamination with air-borne microorganisms.<br />
He showed that putrefaction did not occur in heated broth even with free<br />
access to air, provided that ingress of micro-organisms was prevented by the ingenious<br />
design of his `swan-necked' flasks.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Pasteur made careful microscopic examination of beer fermentations, the results of<br />
which were published in his EÂtudes sur la bieÁre of 1876. He observed the growth of<br />
brewing yeast cells <strong>and</strong> demonstrated that these were responsible for fermentation. He<br />
also recorded the symptoms of several specific `diseases' of fermentation <strong>and</strong> identified<br />
the causative microbial contaminants associated with each. Thus, he was instrumental in<br />
pointing out the necessity of adopting the highest st<strong>and</strong>ards of hygiene for successful<br />
brewing. He designed equipment that fulfilled this requirement. Central to his<br />
recommendations was the realization that microbial growth on sugars or wort was via<br />
contamination <strong>and</strong> not by spontaneous generation. Furthermore, new microbial cells<br />
arose solely from similar parental cells. In order to reach these conclusions Pasteur<br />
developed methods for the sterilization of media <strong>and</strong> the preservation <strong>and</strong> propagation of<br />
microbial cultures.<br />
Pasteur recommended the use of microscopy as an aid in production brewing. Others<br />
have pointed out that this instrument had already been adopted by several brewers prior to<br />
Pasteur's visits (Anderson, 1995). In any case by the early twentieth century, JoÈrgensen in<br />
the third edition of his treatise entitled Micro-organisms <strong>and</strong> Fermentation, published in<br />
English translation in 1900 (JoÈrgensen, 1900) described methods for assessing yeast<br />
`health' by microscopic observation. Pasteur was comparatively indifferent to the<br />
nuances of cellular morphology of individual species. The ability to differentiate microorganisms<br />
was largely dependent on the development of methods for isolating <strong>and</strong><br />
propagating pure cultures. This was difficult using the liquid cultures employed by<br />
Pasteur.<br />
The use of solid microbiological media on which pure cultures could be isolated from<br />
colonies derived from single cells was pioneered by the work of the medical<br />
bacteriologist, Robert Koch (1881). In 1883, Emil Hansen, working at the Carlsberg<br />
Foundation in Copenhagen, used similar techniques to isolate the first pure yeast culture.<br />
At the same time Hansen introduced the first modern apparatus for propagating yeast <strong>and</strong><br />
the first pure culture of `Carlsberg Yeast Number 1' was used successfully in commercial<br />
brewing (Curtis, 1971). The availability of pure cultures provided the means of<br />
elucidating the yeast life cycle. Hitherto, yeast cells had been considered to be merely<br />
phases in the life cycles of other organisms such as moulds, bacteria or even algae (Rose<br />
<strong>and</strong> Harrison, 1971). An assistant of Hansen, SchioÈnning reported the occurrence of a<br />
sexual phase in the yeast life cycle. This was confirmed in the 1930s when éjvind Winge<br />
also working at the Carlsberg Foundation provided a full description of the yeast haplo<strong>and</strong><br />
diplophases.<br />
As early as 1897, BuÈchner demonstrated the formation of ethanol <strong>and</strong> carbon dioxide<br />
from sugar using a cell-free extract of yeast, thereby providing the foundation for the<br />
development of modern biochemistry. Yeast has been used as a convenient experimental<br />
organism in many subsequent investigations. The zymologist, A. H. Rose, proposed in<br />
the introduction to the second volume of the first edition of The Yeasts (Rose <strong>and</strong><br />
Harrison, 1971) the initiation of `Project Y'. This suggested that yeast be used as a model<br />
eukaryotic organism in an integrated approach to the study of cell biology. This challenge<br />
has been taken up <strong>and</strong> the academic literature devoted to yeast in general <strong>and</strong><br />
Saccharomyces cerevisiae in particular is now immense. Many of the discoveries in cell<br />
biology, physiology, biochemistry <strong>and</strong> genetics were made using yeast cells. Probably, S.<br />
cerevisiae is the most extensively studied cell. This has culminated in the sequencing of<br />
the entire genome of S. cerevisiae, the first species for which this has been accomplished<br />
(Goffeau et al., 1996).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
11.2 Taxonomy<br />
Taxonomy is the <strong>science</strong> of the classification of organisms. Using criteria such as<br />
morphology, life cycle, immunological properties, biochemical capabilities <strong>and</strong> genetic<br />
analysis, organisms are grouped into hierarchies of relatedness <strong>and</strong> difference. Systems of<br />
taxonomy indicate functional <strong>and</strong> evolutionary relationships between groups of<br />
organisms <strong>and</strong> they provide a framework for identifying unknown types. Taxonomy<br />
has practical importance in brewing. It allows the identification of proprietary yeast<br />
strains <strong>and</strong> the ability to distinguish these from contaminants such as wild yeasts.<br />
Each group is termed a taxon. In descending order of hierarchy the main taxonomic<br />
groups are kingdom, division, class, order, family, genus, species <strong>and</strong> strain. Of these,<br />
the last three are of most practical interest. A genus represents a group of organisms,<br />
which are closely related in evolutionary terms. Usually this is accompanied by<br />
structural <strong>and</strong> functional similarities. Organisms are grouped into species usually based<br />
on the ability to interbreed. In the case of yeast, many of which have no sexual cycle, this<br />
definition is of limited value. Where organisms are restricted to asexual reproduction,<br />
placement within a species has to be based on other criteria, the most reliable being that<br />
of similarity of genotype. Strains are clones derived from a single parental cell. For<br />
example, in the case of yeast a cell line propagated from a single colony of a pure<br />
culture. In this case, it as assumed that the colony was formed from asexual reproduction<br />
of a single parental cell.<br />
Organisms are described using a binomial nomenclature of genus name followed by<br />
species name. Both are Latin terms <strong>and</strong> by convention in typescript are italicized, the<br />
genus name being capitalized <strong>and</strong> the species name in lower case. After introducing the<br />
complete binomial name, the genus name may be abbreviated thereafter. Thus, brewing<br />
yeasts are classified as Saccharomyces cerevisiae (S. cerevisiae). Bionomial names may<br />
be descriptive of the appearance of the organism, its mode of growth, habitat or a Latin<br />
derivation of the name of the discoverer. The genus name, Saccharomyces translates as<br />
`sugar fungus', referring to the habitat in which the organism is usually to be found <strong>and</strong><br />
the fact that it is a member of the Fungi. The species name cerevisiae derives from the<br />
Latin for beer as in ceres (ˆ grain) <strong>and</strong> vise (ˆ strength).<br />
Genus <strong>and</strong> species names are subject to stringent rules. In the case of yeast, these fall<br />
under the remit of the International Code of Botanical Nomenclature (Greuter et al.,<br />
1996). Strain names are chosen by the discoverer <strong>and</strong> conventions are comparatively lax.<br />
Commonly, they are named using codes, which are combinations of letters <strong>and</strong> numbers.<br />
Many are proprietary strains owned <strong>and</strong> jealously guarded by individual companies. The<br />
ability to identify individual strains <strong>and</strong> maintain them as pure cultures is of the utmost<br />
importance to those who use yeasts in industrial applications. Many thous<strong>and</strong>s of<br />
industrial strains of S. cerevisiae are in existence. In this regard, the finer points of<br />
taxonomy are of somewhat academic interest only. Commonly, systems of nomenclature<br />
are used that are now not recognized by `classical' taxonomists, however, they are<br />
retained because they are still of practical value. For example, see the discussion later in<br />
this section regarding the taxonomy of ale <strong>and</strong> lager brewing strains.<br />
`Yeast' <strong>and</strong> Saccharomyces are not synonymous terms. This is perhaps underst<strong>and</strong>able,<br />
bearing in mind the economic importance of Saccharomyces strains in brewing<br />
<strong>and</strong> baking <strong>and</strong> numerous papers reporting work in which S. cerevisiae has been used as a<br />
type eukaryotic cell. In fact, approximately 100 genera of yeast encompassing 700<br />
species have been described (Kurtzman <strong>and</strong> Fell, 1998). Undoubtedly, many more remain<br />
unrecognized. Thus, some 70,000 species of fungi are currently recognized. It has been<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
estimated that approximately 1.5 million species of fungi may exist (Hawksworth, 1991).<br />
Undoubtedly many of these will be classified within the yeast group.<br />
Kurtzman <strong>and</strong> Fell (1998) define yeasts as being fungi with vegetative states that<br />
reproduce by budding or fission resulting in growth that is predominantly in the form of<br />
single cells. Yeasts do not produce sexual states within or upon aspecialized fruiting<br />
body. This definition is relatively imprecise since many fungi are dimorphic. During<br />
certain phases in their life cycles, such fungi adopt ayeast-like unicellular form <strong>and</strong> at<br />
others they take on afilamentous hyphal habit <strong>and</strong> develop into amycelium.<br />
<strong>Brewing</strong> yeast strains are ascomycetous types classified within the genus<br />
Saccharomyces. The precise taxonomy of the fungi in general <strong>and</strong> the Saccharomyces<br />
in particular is still subject to debate <strong>and</strong> continual revision. Acurrent version is given in<br />
Table11.1.Atpresent,thegenusSaccharomycesisdividedinto14species.Thenamesof<br />
these species together with akey for their differentiation are shown in Table 11.2. The<br />
relationships between the 14 species assigned to the genus Saccharomyces have been the<br />
subject of intensive investigation. Based on homologies between ribosomal RNA <strong>and</strong> the<br />
mitochondrial genome it has been demonstrated that they can be arranged into smaller<br />
sub-groups (Vaughn-Martini <strong>and</strong> Martini, 1993; Piskur et al., 1998; Montrocher et al.,<br />
1998). The Saccharomyces sensuo lato group includes S. dairensis, S. castelli, S. exiguus,<br />
S. servazzii, S. unisporus <strong>and</strong> possibly S. kluyveri. This grouping is considered to contain<br />
species that are relatively heterogeneous.<br />
The Saccharomyces sensuo stricto group contains S. cerevisiae, S. pastorianus, S.<br />
bayanus <strong>and</strong> S. paradoxus. As the nomenclature suggests these are much more closely<br />
related. Based on biochemical differences this group can be sub-divided into two subsets.<br />
S. cerevisiae <strong>and</strong> S. paradoxus can utilise melibiose, assimilate fructose via a<br />
facilitated transport mechanism <strong>and</strong> have a maximum growth temperature close to 37 ëC.<br />
Conversely, S. bayanus <strong>and</strong> S. pastorianus cannot utilize melibiose, have active transport<br />
systems for fructose uptake <strong>and</strong> are not capable of growth above 34 ëC.<br />
All yeast species in the Saccharomyces sensuo stricto group, with the exception of S.<br />
paradoxus, are exploited commercially for the production of ethanol or in baking. It is<br />
suggested that these commercial yeast strains actually arose via selective processes in<br />
industrial situations. S. paradoxus alone is known to occur in natural habitats, whereas the<br />
others are rarely found so (Vaughn-Martini <strong>and</strong> Martini, 1993). S. cerevisiae is used for<br />
baking as well as brewing <strong>and</strong> in winemaking. S. bayanus strains are utilized solely for<br />
enological purposes, whereas S. pastorianus includes those strains originally classified as<br />
S. carlsbergensis <strong>and</strong> used as bottom-fermenting lager yeasts. It is possible, therefore that<br />
S. bayanus arose because of its ability to ferment at the relatively low temperatures of<br />
wine fermentations <strong>and</strong> to withst<strong>and</strong> high ethanol concentrations. Similarly, S.<br />
pastorianus strains were selected for their ability to bottom-crop in low-temperature<br />
lager fermentations.<br />
The phylogenetic relationships between the Saccharomyces sensuo stricto highlight<br />
the undoubted differences between the individual species <strong>and</strong> support the arguments<br />
regarding their origins. Strains of the species S. pastorianus show a high degree of DNA<br />
homology with those of both S. cerevisiae <strong>and</strong> S. bayanus. Conversely, the strains of the<br />
latter two species show little similarity with each other. The genome of S. pastorianus<br />
strains is considerably bigger than that of S. cerevisiae <strong>and</strong> S. bayanus. It is proposed,<br />
therefore, that S. pastorianus is a hybrid species derived from S. cerevisiae <strong>and</strong> either S.<br />
bayanus or a closely related species, S. monacensis (Wolfe <strong>and</strong> Shields, 1997). The latter<br />
is now classified as S. bayanus but was originally considered to be S. pastorianus.<br />
Seemingly, the proportion of DNA contributed by S. bayanus to the S. pastorianus<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.1 Classification of Saccharomyces cerevisiae<br />
Taxon Name Comments<br />
Kingdom Fungi<br />
Phylum Ascomycotina Teliomorphic forms characterized by formation of ascospores enclosed within ascus<br />
Sub-phylum Saccharomycotina (syn. Hemiascomycotina)<br />
Class Saccharomycetes (syn. Hemiascomycetes) Single ascus not enclosed in ascocarp developing directly from zygotes<br />
Order Saccharomycetales (syn. Endomycetales) Yeast-like cells, rarely developing hyphae<br />
Family Saccharomycetaceae<br />
Genus Saccharomyces Globose, ellipsoidal or cylindroidal cells. Vegetative reproduction by multilateral budding.<br />
Pseudohyphae may be formed but hyphae are not septate. The vegetative form is predominantly<br />
diploid, or of higher ploidy. Diploid ascopores may be formed that are globose to short ellipsoidal<br />
with a smooth wall. There are usually 1±4 ascopores per ascus<br />
Type species S. cerevisiae<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.2 Key to the species of Saccharomyces (taken from Kurtzman <strong>and</strong> Fell, 1998)<br />
1 a. Maximum growth temperature above 30 ëC ! 3<br />
b. Growth absent above 30 ëC ! 2<br />
2 a. Sucrose, raffinose <strong>and</strong> trehalose fermented S. barnettii<br />
b. Sucrose, raffinose <strong>and</strong> trehalose not fermented S. rosini<br />
3 a. Ethylamine-HCl assimilated ! 4<br />
b. Ethylamine-HCl not assimilated ! 6<br />
4 a. Growth in the presence of 1000 ppm cycloheximide S. unisporus<br />
b. No growth in the presence of 1000 ppm cycloheximide ! 5<br />
5 a. Maltose, raffinose <strong>and</strong> ethanol assimilated S. kluyveri<br />
b. Maltose, raffinose <strong>and</strong> ethanol not assimilated S. spencerorum<br />
6 a. Maltose assimilated ! 7<br />
b. Maltose not assimilated ! 10<br />
7 a. Growth in vitamin-free medium S. bayanus<br />
b. No growth in vitamin-free medium ! 8<br />
8 a. D-mannitol assimilated, maximum growth temperature 37 ëC or greater S. paradoxus<br />
b. D-mannitol not assimilated, maximum growth temperature less than ! 9<br />
37 ëC or variable at 37 ëC<br />
9 a. Active transport mechanism for fructose present; maximum growth S. pastorianus<br />
temperature 34 ëC or lower<br />
b. Active transport mechanism for fructose not present; maximum S. cerevisiae<br />
growth temperature variable<br />
10 a. Sucrose, raffinose <strong>and</strong> trehalose fermented S. exiguus<br />
b. Sucrose, raffinose <strong>and</strong> trehalose not fermented ! 11<br />
11 a. Growth in the presence of 1000 ppm cycloheximide S. servazzii<br />
b. No growth in the presence of 1000 ppm cycloheximide ! 12<br />
12 a. D-ribose normally assimilated; 8±10 chromosomes 600±3000 kilobases S. castellii<br />
b. D-ribose not assimilated, mostly single highly refringent ascospores S. transvaalensis<br />
on acetate agar; 8 chromosomes 400±2200 kilobases<br />
c. D-ribose normally not assimilated, 7±9 chromosomes 750±3000 kilobases S. dairiensis<br />
hybrids is greater than that made by S. cerevisiae. Nevertheless, chromosomes identical<br />
to those from both parents have been found co-existing in strains of S. pastorianus<br />
(Tamai et al., 1998).<br />
The taxonomic history of lager brewing yeast strains is chequered. The original<br />
descriptor for lager strains, S. carlsbergensis changed when these were re-christened as S.<br />
uvarum. Later the position became more clouded when the latter were assigned to the<br />
species S. cerevisiae. As discussed already, bottom-fermenting lager yeasts are again<br />
considered distinct from S. cerevisiae brewing strains <strong>and</strong> have been given the name S.<br />
pastorianus. This confirms what brewers have already decided, based on experience <strong>and</strong><br />
observation, that ale <strong>and</strong> lager yeasts are different (Quain, 1986).<br />
11.3 Yeast ecology<br />
Yeasts are predominantly saprophytes <strong>and</strong> are widely distributed in nature where they are<br />
found in both terrestrial <strong>and</strong> aquatic habitats (Phaff <strong>and</strong> Starmer, 1987). In nature, yeasts<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
are primarily associated with higher plants, although the effects of rain <strong>and</strong> plant death<br />
inevitably means that soils act as areservoir in which they can survive <strong>and</strong> be passed on<br />
to other hosts. Yeasts are rarely plant pathogens, instead they are commonly found on<br />
damaged fruits, in flowers <strong>and</strong> in exudates associated with wounds. The transfer of yeasts<br />
between plants is most often accomplished by the intermediary of insect vectors.<br />
Natural populations of yeasts co-exist <strong>and</strong> compete with themselves <strong>and</strong> with other<br />
microbial species. Many yeast species are found in specialized plant habitats, which<br />
reflect their biochemical capabilities. Contrary to expectation, non-fermentative<br />
obligately aerobic yeast types are the most common. Typically, they occupy niches<br />
that provide a particular oxidizable substrate that they are capable of assimilating.<br />
Fermentativeyeastsareabletotakeadvantageofhabitatswherethereisasourceofsugar<br />
but no oxygen. Since such yeasts are facultative anaerobes the result of their own<br />
metabolic activity would be to remove oxygen from aerobic environments. They would<br />
then be able to continue to grow under conditions of anaerobiosis, whereas purely<br />
oxidative yeasts could not. In aquatic habitats containing asource of fermentable sugar<br />
the result would be that aerobic yeast would be restricted to the surface layers, possibly<br />
resulting in the formation of apellicle. The population of fermentative yeasts would be<br />
capable of growth throughout the body of the liquid.<br />
These observations are paralleled in brewing, for example, the growth of brewing<br />
yeast in afermenter <strong>and</strong> the effects of some spoilage organisms in beer exposed to air.<br />
Some of the associations of yeasts <strong>and</strong> other micro-organisms have elements of<br />
symbiosis. For example, acetic acid bacteria are often found growing on the ethanol<br />
produced as aresult of the metabolism of sugars by yeast. The bacteria remove ethanol,<br />
which can be toxic or even fatal to the yeast. The resultant acetic acid reduces the pH of<br />
the medium <strong>and</strong> inhibits the growth of other less acid-tolerant species. Some yeasts are<br />
able to occupy particular niches because of an ability to tolerate otherwise toxic products<br />
ofthegrowthofothermicro-organisms.Forexample,naturallyoccurringantibioticssuch<br />
as cycloheximide produced by Streptomyces griseus, inhibit the growth of many yeast<br />
species but not others.<br />
Other yeast types produce metabolites that are toxic to other potential yeast<br />
competitors. Some strains of Saccharomyces cerevisie <strong>and</strong> representatives of the genera<br />
C<strong>and</strong>ida, Crytpococcus, Debaromyces, Hansenula, Kluyveromyces, Pichia <strong>and</strong> Torulopsis<br />
produce so-called killer factors, which are fatal to other susceptible yeast species<br />
(Young, 1987). The killer factors, several distinct types of which are produced by<br />
individual yeast genera, are protein or glycoprotein in nature <strong>and</strong> are coded for either by<br />
chromosomal genes or in some cases by the RNA genomes of mycoviruses. They appear<br />
to bind to chitin in the wall of susceptible cells <strong>and</strong> cause death by destroying the<br />
transmembrane potential.<br />
Yeasts are common contaminants of fruits <strong>and</strong> are potential spoilage organisms in<br />
extracted fruit juices, pureÂes <strong>and</strong> concentrates. For example, in one study (Arias et al.,<br />
2002) using orange juices some 99 different yeast strains, representing 11 genera were<br />
isolated. Some yeast species are capable of growth in media with very low water activity.<br />
These so-called osmotolerant species (Section 12.3.1) can cause spoilage of products<br />
such as bulk sugar syrups, particularly if the storage conditions allow condensation to<br />
form on the surface of the liquid.<br />
Yeast strains used in industrial fermentations were originally isolated from nature. In<br />
rare cases, the natural microbial flora of the building in which the fermentation is<br />
conducted is used as the source of the inoculum, for example, in the production of<br />
traditional Belgian Lambic beer. In the modern brewery process, great care is taken to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ensure that contamination of worts <strong>and</strong> beers with foreign `wild' yeast does not occur.<br />
Production strains of brewing yeasts are maintained as pure cultures <strong>and</strong> introduced into<br />
the brewery via asystem of propagation. Such strains have been selected based on their<br />
possession of particular sets of desirable properties. In consequence, as discussed in the<br />
previous section of this chapter, brewing strains appear to be distinct from those found in<br />
nature.<br />
11.4 Cellular composition<br />
Yeast cells contain approximately 80% water. Thus, inpressed brewers'yeast the ratio of<br />
wet to dry weight is roughly 5:1. Predictably, the most abundant element is carbon,<br />
which accounts for just under 50% of the dry weight. Other major elemental components<br />
are oxygen (30±35%), nitrogen (5%), hydrogen (5%) <strong>and</strong> phosphorus (1%). The total<br />
mineral content of yeast is approximately 5±10% of the cell dry weight. This fraction<br />
comprises amultitude of trace elements. The composition of some of these for three<br />
brewing <strong>and</strong> one bakers' yeast strain is shown in Table 11.3. The most abundant classes<br />
of macromolecules are proteins (40±45% cell dry weight), carbohydrates (30±35%),<br />
nucleic acids (6±8%) <strong>and</strong> lipids (4±5%). The precise composition of each class of<br />
macromolecules within agiven cell varies as afunction of physiological condition <strong>and</strong><br />
phase in growth cycle. There is considerable variation between different yeast species.<br />
For these reasons, it is not possible to provide other than the generalized composition<br />
given above. The pathways resulting in the formation <strong>and</strong> dissimilation of the major<br />
classes of macromolecules are discussed in Chapter 12.<br />
Table 11.3 Trace elemental composition of yeast ( g dry wt. yeast)<br />
Bakers Brewers Brewers Brewers<br />
(Reed <strong>and</strong> Nagodarithana, 1991) (Eddy, 1958) (Eddy, 1958) (Eddy, 1958)<br />
Aluminium ± 3.0 2.0 1.0<br />
Calcium 0.75 ± ± ±<br />
Chromium 2.2 37.0 104.0 34.0<br />
Copper 8.0 ± ± ±<br />
Iron 20.0 17.0 104.0 25.0<br />
Lead ± 2.0 14.0 100.0<br />
Lithium 0.17 ± ± ±<br />
Magnesium 1.65 ± ± ±<br />
Manganese 8.0 4.0 5.0 11.0<br />
Molybdenum 0.04 0.1 0.04 2.7<br />
Nickel 3.0 3.0 4.0 3.0<br />
Phosphorus 13.0 ± ± ±<br />
Potassium 21.0 ± ± ±<br />
Selenium 5.0 ± ± ±<br />
Silicon 30.0 ± ± ±<br />
Sodium 0.12 ± ± ±<br />
Sulphur 3.9 ± ± ±<br />
Tin 3.0 3.0 > 100 3.0<br />
Vanadium 0.04 ± ± ±<br />
Zinc 170.0 ± ± ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
11.5 Yeast morphology<br />
Individual yeast cells are not visible to the human naked eye <strong>and</strong> they become evident<br />
only when proliferation produces amass of many millions of cells. When this occurs,<br />
yeastcellstakeontheappearanceofsurfacepellicles,sedimentsorhazesonorwithinthe<br />
bodyofliquids.Saccharomyces yeast enmasse,isoff-white,grey orbeigeinappearance.<br />
Commonly, yeast biomass is stained by components of the growth medium adhering to<br />
theyeast cellwall.<strong>Brewing</strong>yeaststrainsthatarenon-flocculentincharacterformsmooth<br />
creamy slurries in beer. Such yeasts are termed `powdery'. Conversely, flocculent strains<br />
formslurrieswithadistinctgranularappearanceinwhichtheyeastreadilyseparatesfrom<br />
the beer.<br />
When ayeast cellistransferred toalaboratory nutrient mediumsolidifiedwith agar or<br />
gelatin subsequent proliferation results in the formation of aroughly circular mass of<br />
cells termed acolony.Thesize <strong>and</strong>shape of coloniesvaries with the yeastspecies, nature<br />
of the growth medium, the solidification agent <strong>and</strong> the conditions under which the plates<br />
are incubated. Providing these parameters are defined some individual yeast strains give<br />
rise to colonies that have acharacteristic morphology. This has been used as amethod of<br />
yeast differentiation (Section 13.9).<br />
The size <strong>and</strong> shape of cells <strong>and</strong> the patterns of vegetative propagation are<br />
characteristic of individual yeast species <strong>and</strong> may be used as aids to identification<br />
(Fig. 11.1). A description of cells of S. cerevisiae provided by Lodder (1970) is<br />
`spheroidal, subglobose, ovoid, ellipsoidal or cyclindrical to elongate, occurring singly or<br />
in pairs, occasionally in short chains or clusters'. The ratio of the long <strong>and</strong> short axes of<br />
cells of S. cerevisiae averages approximately 1.4:1, although some cells are longer <strong>and</strong><br />
thinner than this. Most fall within the range of 2.5±4.5 m <strong>and</strong> 10.5±20 m along the<br />
short <strong>and</strong> long axes, respectively. Cell volumes range between 50 <strong>and</strong> 500 m 3 .<br />
Cell size is a characteristic of individual strains although there is considerable<br />
variation within strains depending on the phase of the growth cycle <strong>and</strong> cultural<br />
conditions. Thus, the mean cell size increases with increase in incubation temperature,<br />
within the normal range for any given strain. The cell wall of yeast is relatively flexible.<br />
This accommodates transient variations in cell volume in response to sudden shifts in<br />
osmotic pressure of the suspending medium. For example, when yeast slurries suspended<br />
in beer are transferred to wort there is an increase in cell size, which persists for the first<br />
few hours <strong>and</strong> precedes the onset of budding (Quain, 1988). As cells proceed through the<br />
growth cycle, there are changes in cell size <strong>and</strong> density. During the budding phase, the<br />
cell volume decreases by approximately a third. In the intervals between budding, there is<br />
an increase in cell density due to a loss of water (Baldwin <strong>and</strong> Kubitschek, 1984). Cell<br />
size of brewing yeast has been reported to decrease during the period when yeast is stored<br />
between cropping <strong>and</strong> re-pitching (Cahill et al., 1996). During 14 days storage at 4 ëC, the<br />
mean cell volume of an ale yeast fell from 302 to 244 m 3 <strong>and</strong> a lager strain from 208 to<br />
194 m 3 , ascribed to reduction in biomass due to glycogen turnover.<br />
The mean cell size increases with generational age. For any given strain there is a<br />
rough correlation between number of bud scars <strong>and</strong> mean cell size. Using a diploid lager<br />
yeast strain Barker <strong>and</strong> Smart (1996) reported that the mean cell volume of virgin<br />
daughter cells was approximately 150 m 3 . Cells, which had undergone around 20 rounds<br />
of budding <strong>and</strong> were approaching the end of their life spans, had mean cell volumes of<br />
approximately 850 m 3 . <strong>Brewing</strong> yeast cells tend to be bigger than haploid laboratory<br />
strains of S. cerevisiae. This is a consequence of the fact that the former tend to be<br />
polyploid/aneuploid. Galitski et al., (1999) measured the mean cell volume of haploid,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a) (b) (c)<br />
(d) (e) (f)<br />
Fig. 11.1 Drawings of vegetative cells of (a) Saccharomyces cerevisiae (multilateral budding);<br />
(b) Schizosaccharomyces pombe (vegetative reproduction by binary fission); (c) Nadsonia sp.<br />
(apiculate (lemon) shaped cells undergoing bipolar budding); (d) Sterigmatomyces halophilus<br />
(conidiospores borne on short stalks); (e) Trigonopsis variablis (triangular shaped cells buds borne<br />
in angles); (f) Oosporidium margaritiferum (multilateral budding in which cells remain attached<br />
<strong>and</strong> form long chains).<br />
diploid,triploid<strong>and</strong>tertraploidconstructsofastrainofS.cerevisiae.Thesewerefoundto<br />
be 72, 111, 152 <strong>and</strong> 289 m 3 ,respectively.<br />
11.6 Yeast cytology<br />
Yeast cell ultrastructure became amenable to detailed study with the advent of electron<br />
microscopy. Such techniques include scanning electron microscopy, which allows<br />
detailed examination of surface topology. Amore recent development, atomic force<br />
microscopy, performs asimilar function although in this case at the level of individual<br />
molecules (De Souza et al., 1996). Confocal microscopy provides three-dimensional<br />
imaging of cells by measuring fluorescence light intensity produced by alaser-scanning<br />
device (Bacallao <strong>and</strong> Stelzer, 1989). Transmission electron microscopy is suitable for<br />
producing images of very thin sections of cells. The study of individual cellular<br />
organelles was largely dependent on the development of techniques that allowed the<br />
controlled disruption of yeast cells <strong>and</strong> recovery of undamaged organelles (Lloyd <strong>and</strong><br />
Cartledge, 1991).<br />
An idealized representation of asection through abudding yeast cell is shown in Fig.<br />
11.2. The diagram shows all the major organelles present that may occur in the cell. Not<br />
all of these are visible at any given time. In some non-brewing yeast genera, there is an<br />
external capsule, which is attached to the outside of the wall. The cell wall, plus capsule<br />
if present, plasma membrane <strong>and</strong> the intervening periplasmic space are collectively<br />
referred to as the cell envelope. Modifications to some organelles occur as the cell<br />
progresses through the cell cycle. Some changes occur in response to growth conditions.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Plasma membrane<br />
Cytoplasm<br />
Tonoplast<br />
Ribosomes<br />
Volutin granule<br />
Mitochondrion<br />
Cell wall<br />
Endoplasmic reticulum<br />
Bud scar<br />
Vacuole<br />
Nucleolus<br />
Nucleus<br />
Mitochondrion<br />
migrating in<br />
emerging bud<br />
Periplast<br />
Golgi body<br />
Lipid granule<br />
Glycogen granule<br />
Birth scar<br />
Fig. 11.2 Diagrammatic representation of asection through atypical budding yeast cell.<br />
11.6.1 Cell wall<br />
The wall is the outermost layer of the cell. It is of rugged construction, typically between<br />
150 <strong>and</strong> 200nm in thickness <strong>and</strong> accounts for approximately 20% of the total cell dry<br />
weight (Smits et al., 1999). Cells that have undergone vegetative reproduction by<br />
budding bear characteristic circular bud scars. These mark the point on the wall at which<br />
the daughter cell was excised from the mother. The bud scar region is relatively rich in<br />
chitin.Thiscan beseen followingtreatmentwithfluorescent dyes suchascalcofluor.The<br />
point on the daughter cell wall, which corresponds to the bud scar on the mother cell, is<br />
termed the birth scar. Buds do not arise r<strong>and</strong>omly on the cell surface but occur at specific<br />
locations. The patterns of bud scars are frequently characteristic of individual species.<br />
The process of budding is described in more detail in Section 11.7.<br />
The cell wall is approximately 90% carbohydrate, the remainder being protein. A<br />
diagrammatic representation of asection through the cell wall is shown in Figs 11.3±4.<br />
The most abundant carbohydrates are glucans, which make up around 30±50% of the<br />
total dry weight of the wall. Glucans are arranged into long fibrillar structures, which are<br />
joined together by -1,3 <strong>and</strong> -1,6 linkages. Most of the remaining cell wall<br />
carbohydrate is mannoprotein (Fig. 11.5). This is comprised of an inner core region of<br />
repeating -1,6linked mannose (4.8) residues with short side chains attached via -1,2<br />
<strong>and</strong> -1,3linkages. The inner core region is attached to an outer chain of 100±150<br />
mannose residues.Thisalsoconsistsofabackboneof -1,6linked mannose residue with<br />
side chains attached via -1,2linkages. The side chains are of mannobiose (M 2 1 M),<br />
mannotriose (M 2 1 M 2 1 M), mannotriose (M 2 1 M 3 1 M) <strong>and</strong> mannotetraose<br />
(M 2 1 M 2 1 M 3 1 M). The precise composition of these side chains varies between<br />
yeaststrains.Someofthesidechainscontainphosphodiesterlinkages<strong>and</strong>theseconferan<br />
overall negative charge to the cell envelope. The side chains of the mannose molecules<br />
are the sites of the receptors, which are implicated in yeast flocculation (Section<br />
11.6.1.1). The other end of the inner core region is attached to two N-acetyl glucosamine<br />
residues. One of these is attached to aprotein molecule via the carboxylic acid moiety of<br />
an aspartic acid residue. Attached to the protein via the hydroxyl groups of serine (4.45)<br />
<strong>and</strong> threonine (4.46) are short -1,2<strong>and</strong> -1,3chains of mannose residues.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
EXTERIOR<br />
Glycogen<br />
INTERIOR<br />
Chitin cross-linked<br />
to β-glucans<br />
β,1-3 glucans<br />
Mannoproteins<br />
covalently linked<br />
to glucans<br />
Mannoproteins<br />
anchored in<br />
membrane<br />
β,1-3 <strong>and</strong> β,1-6<br />
glucans<br />
Periplasmic space<br />
Periplasmic enzyme<br />
Plasma membrane<br />
Fig. 11.3 Diagrammatic representation of a section through the yeast cell envelope.<br />
(a)<br />
CH2OH<br />
H<br />
O<br />
H<br />
H<br />
OH HO<br />
HO OH<br />
H H<br />
H<br />
H<br />
HO OH<br />
NH·CO·CH3<br />
Chitin consists of a linear polymer of molecules of N-acetyl glucosamine linked by -<br />
1, 4 groups. It accounts for less than 5% of the dry weight of the wall in S. cerevisiae.<br />
Almost all is located within the bud scars although a small quantity is distributed<br />
throughout the rest of the wall. The fourth <strong>and</strong> also minor carbohydrate component of cell<br />
walls is glycogen. It is acid soluble <strong>and</strong> distinct from the alkali soluble pool, which<br />
functions as a reserve material. It has been demonstrated that the acid soluble fraction is<br />
structural <strong>and</strong> is linked to -, 1 3 glucans via the -1, 6 glucan side-chains (Arvindekar<br />
<strong>and</strong> Narayan, 2002). Most of the protein component of the wall is associated with<br />
mannose. This fraction confers immunological properties to the cell. Some of the proteins<br />
are surface receptors <strong>and</strong> others are enzymes. The latter are those responsible for cell wall<br />
biosynthesis <strong>and</strong> the initial metabolism of some nutrients (Fleet, 1991).<br />
The precise macromolecular structure of the cell wall remains uncertain. The glucan<br />
fibres are mainly located within the inner part of the wall. It is considered that the fraction<br />
of glucans attached by -1, 3 linkages form an interwoven network of fibrils responsible<br />
for conferring strength <strong>and</strong> flexibility to the wall. The -1, 6 linked glucans form a<br />
(b)<br />
CH2OH<br />
H<br />
O OH<br />
H H<br />
(11.1) Mannose, α-D-mannopyranose (11.2) β-D-N-Acetyl glucosamine<br />
Fig. 11.4 The structures of (a) -D-mannose <strong>and</strong> (b) -D-N-acetyl glucosamine.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Asn<br />
Ser<br />
or ( thr )<br />
Ser<br />
or ( thr )<br />
Ser<br />
or ( thr )<br />
Ser<br />
or ( thr )<br />
1 4<br />
Gln Ac<br />
Protein<br />
1 M<br />
1 M 2 1 M<br />
1 M 2 1 M 2<br />
1 M<br />
M 1 2<br />
M 1 3<br />
M 1<br />
1 M 2 1 M 2 1 M 2 1 M 2<br />
Protein<br />
1 4<br />
Gln Ac<br />
1 6<br />
M3<br />
1 6<br />
M2<br />
M 1 3<br />
M 1<br />
M3 M2 M<br />
M 1 M 1<br />
(M2 M2 M2 M2)n<br />
P M 1 2 M 1 2 M 1 2<br />
M 1 3 M 1 3 M 1 3<br />
M 1 M 1 M 1<br />
Inner core Outer chain iso mannoses<br />
Fig. 11.5 Yeast phospho-mannoprotein. M. mannose; Gln-Ac, N-acetyl glucosamine; asn,<br />
asparagine; ser, serine; thr, threonine.<br />
connection between this fibrillar network <strong>and</strong> the mannoprotein, glycogen <strong>and</strong> chitin<br />
components. The mannoproteins are mainly situated towards the outside of the wall<br />
where they form a cross-linked layer covalently bonded to the glucans. The extent of<br />
cross-linking between the mannoproteins appears to determine the size of molecule that<br />
can pass through the wall. A second class of mannoproteins is attached to the plasma<br />
membrane <strong>and</strong> project across the periplasmic space through the glucan layer (Fleet,<br />
1991). These mannoproteins are implicated in flocculation <strong>and</strong> sexual agglutination. The<br />
fraction of chitin not located in bud scars is distributed throughout the body of the cell<br />
wall. Its function is unclear although it appears to have structural significance since<br />
mutants lacking it are sensitive to osmotic shock (Fleet, 1991). Chitin is the receptor for<br />
binding of killer toxin to yeast cells (Takita <strong>and</strong> Castilho-Valavacius, 1993).<br />
The cell wall has several functions. It forms a protective layer over the comparatively<br />
fragile plasma membrane. It has a degree of flexibility, which allows rapid fluctuations in<br />
cell volume in response to changes in the osmotic potential of the external medium.<br />
Conversely, it has sufficient mechanical strength to prevent lysis when cells are subject to<br />
hypo-osmotic shock. This rigidity is responsible for conferring characteristic shapes to<br />
individual cells. The generalized rigidity <strong>and</strong> targeted weakening, together with the<br />
motive force of turgor pressure provides the means for bud development. The cell wall is<br />
a repository for several enzymes <strong>and</strong> it ultimately limits the size of molecules that may<br />
pass into <strong>and</strong> out of the periplasm. The cell wall is important in determining interactions<br />
between cells <strong>and</strong> with the external medium.<br />
During fermentation some wort components, notably trub lipids <strong>and</strong> hop iso- -acids,<br />
bind to yeast cell walls. Hop iso- -acids dosage rates must be adjusted to allow for the<br />
proportion lost with the yeast crop. The impact of binding of trub components to yeast<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
M 1<br />
M 1
cell walls is more difficult to assess. Potentially, the bound wort components will block<br />
receptor sites implicated in flocculation. Possibly, binding of trub components will<br />
change the overall cell surface charge <strong>and</strong> by implication affect fining behaviour.<br />
Isinglass finings are added to green beer to encourage yeast cells to aggregate together<br />
<strong>and</strong> forms flocs, which promotes sedimentation. The active component in isinglass is the<br />
positively charged protein collagen, which binds electrostatically to negatively charged<br />
yeast cells. Pre-loading of yeast cells with positively charged material interferes with the<br />
action of isinglass (Leather et al., 1997).<br />
The cell wall plays a role in determining whether or not yeast rises to the surface (top<br />
yeast) or sediments (bottom yeast) during wort fermentation. Differences in the two types<br />
of yeast can be demonstrated simply by placing a suspension of yeast in water in a test tube<br />
<strong>and</strong> shaking the mixture. Top yeasts form a surface pellicle at the interface between water<br />
<strong>and</strong> air. Bottom yeasts remain distributed throughout the water. Top fermenting yeasts form<br />
loose flocs during fermentation. These trap rising carbon dioxide bubbles <strong>and</strong> the mass of<br />
gas <strong>and</strong> yeast form a type of foam akin to a head on a beer. The cell walls of top fermenters<br />
are more hydrophobic than bottom yeasts. It is been demonstrated that hydrophobicity<br />
shows a negative correlation with the content of phosphate in the outer part of the cell wall<br />
(Mestdagh et al., 1990). Highly hydrophobic cells (low cell wall phosphate content) would<br />
tend to be attracted to bubbles <strong>and</strong> the surface of polar liquids such as fermenting wort.<br />
The phosphate content of the wall is the major determinant of another measure of the<br />
cell surface, that is zeta potential (Lawrence et al., 1989). This parameter is a measure of<br />
the surface charge of yeast cells. Apart from the phosphate content of the wall, it is also<br />
influenced by the pH of the surrounding medium. Zeta potential reportedly declines<br />
(becomes less negative) either during or at the end of fermentation (Iserentant, 1996).<br />
This would tend to reduce the electrostatic repulsion between individual cells <strong>and</strong> thereby<br />
favour the formation of flocs.<br />
11.6.1.1 Flocculation<br />
Flocculation is the reversible process by which some yeast cells adhere to each other to<br />
form aggregates. It is distinct from aggregates, which arise via budding <strong>and</strong> nonseparation<br />
of daughter cells. Flocculation is of enormous significance to brewing. The<br />
propensity of yeast to form flocs is an integral part of the process of separating the crop<br />
from green (immature) beer. Top fermenting types form flocs that rise to the surface of<br />
the fermenting vessel. The resultant yeast head can be removed by skimming or suction.<br />
Bottom fermenting yeast form flocs which settle into the base of the fermenting vessel, a<br />
process which is encouraged by chilling the green beer. The bottom crop can be removed<br />
from the fermenter before the beer is racked. The formation of flocs is an essential<br />
precursor of crop formation. Inadequate flocculation results in poor cropping such that<br />
there may be insufficient yeast for re-pitching <strong>and</strong> green beer with unacceptably high<br />
residual yeast counts. Conversely, if flocculation occurs too soon, fermentation may<br />
arrest because insufficient cells remain suspended in the fermenting wort.<br />
Flocculation is observed in strains from several genera. Conversely many strains,<br />
including several Saccharomyces brewing strains, do not flocculate to any great extent<br />
under any circumstances. <strong>Brewing</strong> strains possessing desirable flocculation characteristics<br />
will have been chosen by `natural selection' as being the most suitable for use with<br />
particular combinations of wort <strong>and</strong> fermenting vessels. Flocculation <strong>and</strong> flocculence are<br />
distinct. The latter is an inherent property that some strains possess. The former refers to<br />
the expression of flocculence in those strains capable of so doing. By inference,<br />
flocculation is not expressed under all conditions. Commonly, flocculation occurs only<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
when sources of fermentable sugars are exhausted. It has been suggested (Iserentant,<br />
1996)thatundersuchstarvationconditionstheabilitytoformflocsmayrepresentastress<br />
response. Thus,flocs provide asheltered environment where thechanceofsurvival ofthe<br />
population is enhanced. Disaggregation of flocs occurs if the cells are again exposed to a<br />
source of fermentable sugar. In this case, the re-adoption of asingle cell mode affords<br />
unimpeded opportunity to utilize the supply of sugar.<br />
The precise mechanism by which flocculation occurs is controversial <strong>and</strong> there is no<br />
consensus that there is asingle mechanism that applies to all yeast strains. The onset of<br />
flocculation is observed in laboratory cultures that have just entered the stationary phase<br />
of growth. Similarly, in brewing, flocculation occurs towards the end of primary<br />
fermentation. Nevertheless, exponentially growing yeast cells can be made to flocculate<br />
providing they are removed from the growth medium <strong>and</strong> washed <strong>and</strong> suspended in<br />
water, supplemented with Ca 2+ ions. Cells must come into contact with each other for<br />
flocculation to occur, hence the surprising observation that flocculation <strong>and</strong> the vigour of<br />
mechanical agitation are positively correlated. Thus, in well-stirred systems there is a<br />
high probability that cells will contact each other <strong>and</strong> once formed, flocs are relatively<br />
stable structures.<br />
Flocculation occurs because of interactions between surface proteins on one cell <strong>and</strong><br />
carbohydrate receptors on another cell (Miki et al., 1982). It has been demonstrated that<br />
flocculation can be inhibited by pre-treatment of cells with agents that block these<br />
interactions. This has allowed classification of yeasts based on the nature of agents that<br />
inhibit flocculation. The evidence suggests that the protein component is alectin since an<br />
excess of related lectins, such as Concanavilin A, abolishes flocculation. Similarly,<br />
simplesugars,whichalsobindtolectin-likeproteins,alsopreventorreverseflocculation.<br />
Three phenotypes have been recognized based on which sugars inhibit flocculation<br />
(Stratford <strong>and</strong> Assinder, 1991; Dengis et al., 1995). Flo1 types do not flocculate in the<br />
presence of mannose, whereas mannose, sucrose, glucose <strong>and</strong> maltose abolish<br />
flocculation of NewFlo types. The MI phenotype flocculates in the presence of both<br />
mannose <strong>and</strong> sucrose but not in the absence of ethanol (Table 11.4). The MI phenotype is<br />
totally distinct from the other two groupings. In these cells flocculation occurs via direct<br />
(non-lectin like) protein ±protein interaction. These strains are top-fermenters <strong>and</strong> have<br />
highly hydrophobic cell envelopes. Possibly the latter promotes both the formation of<br />
flocs <strong>and</strong> encourages formation of ayeast head.<br />
Flo1 <strong>and</strong> NewFlo types all use interactions between lectin-like proteins <strong>and</strong> cell<br />
surface mannans. The groups differ in the nature of the lectins. These can be<br />
differentiated based on differences in patterns of proteolytic digestion <strong>and</strong> response to<br />
pH. Synthesis of particular lectins is dependent on the possession of the relevant genes.<br />
Geneticdifferencesunderpintheflocculationphenotypicclassification.BothNewFlo<strong>and</strong><br />
Flo1 phenotypes use common carbohydrate receptors. Based on studies with mutants<br />
Stratford (1992) has demonstrated that the receptors are the side chains of the outer<br />
mannose chain of cell wall mannoproteins (Fig. 11.5). NewFlo <strong>and</strong> Flo1 types have an<br />
obligate requirement for Ca 2+ ions for flocculation to occur. Absence of this ion, or the<br />
presence of chelating agents prevent flocculation. The role of calcium is probably that of<br />
ensuring that the lectin-like protein is in the correct configuration for binding to the<br />
mannose receptors.<br />
The stability of flocs is proportional to the number of interactions between individual<br />
cells <strong>and</strong> the number of potential binding sites, both protein <strong>and</strong> mannose per unit area of<br />
cell wall will be influential. In addition, spatial considerations must play a role in<br />
allowing cells to pack together <strong>and</strong> make the intimate contacts necessary for interactions<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.4 Classification of yeast flocculence phenotypes (Straford, 1994)<br />
Class Phenotype Inhibitors Requirement Comments<br />
for Ca 2+<br />
Flo 1 Heavily flocculent Mannose Yes Mannose absent from<br />
throughout fermentation wort<br />
New Flo Most brewing strains Mannose, sucrose Yes Flocculation at end of<br />
glucose, maltose primary fermentation<br />
MI Heavily flocculent <strong>and</strong> Flocculation not No Cells require presence<br />
top-fermenting inhibited by sugars of ethanol for<br />
flocculation to occur<br />
tooccur.Thissuggeststhatelectrostatic<strong>and</strong>hydrophobiceffectsmightalsobeimportant.<br />
In brewing, the onset of flocculation observed at the end of primary fermentation is<br />
explainedbecausetheinhibitoryeffectofsugarsisalleviatedbytheirexhaustionfromthe<br />
wort. Similarly, cells disaggregate at the start of fermentation when yeast suspended in<br />
beer is pitched into fresh sugar-containing wort.<br />
The genetics of flocculation is complex <strong>and</strong> still not fully characterized (for areview<br />
see Jin <strong>and</strong> Speers, 1998). Anumber of genes have been identified, termed FLO, which<br />
are reportedly implicated in flocculation. It is assumed that some of these encode for the<br />
lectin-like proteins. The possession of genes producing different lectin-like proteins<br />
presumably underpins the NewFlo <strong>and</strong> Flo1 phenotypes. Strains that do not possess any<br />
of these genes are not flocculent under any circumstances. Thus, there is evidence that a<br />
gene, termed FLO1 encodes for acell surface protein, which has been implicated in<br />
flocculation. Transfer of this gene from aflocculent yeast strain to anon-floccculent type<br />
is accompanied by the acquisition of aflocculent phenotype (Teunissen <strong>and</strong> Steensma,<br />
1995).<br />
11.6.2 The periplasm<br />
Theperiplasm is thespace betweenthe cellwall <strong>and</strong> theplasma membrane. Although not<br />
an organelle as such it is avoid that must be traversed by all in-coming nutrients <strong>and</strong> outgoing<br />
metabolic by-products. It is the location of several enzymes including invertase,<br />
acid phosphatase, melibiase <strong>and</strong> various binding proteins. These are produced<br />
extracellularly, that is exterior to the plasma membrane, but they are retained by virtue<br />
of being too large to pass through the cell wall. The periplasm is not acontinuous void<br />
sincesomecomponents ofthe cellwall areanchoredintheplasma membrane (Fig.11.3).<br />
Possibly (Arnold, 1991) the retention of enzymes that are capable of hydrolysing<br />
otherwise non-assimilable nutrients affords a competitive advantage compared to<br />
organisms that excrete free extracellular enzymes. The protein content of the periplasm<br />
may be sufficiently high to make the fluid gel-like. Thus, the periplasm may also function<br />
as a protective layer between the cell wall <strong>and</strong> plasma membrane.<br />
11.6.3 The plasma membrane<br />
Plasma membranes enclose the cytoplasm <strong>and</strong> form the inner barrier between the cell<br />
wall <strong>and</strong> periplasm. The yeast plasma membrane resembles that of other eukaryotic cells.<br />
It consists of roughly equal quantities of lipid <strong>and</strong> protein. The proteins are diverse <strong>and</strong><br />
mainly functional as opposed to structural. They include the enzymes responsible for cell<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
wall synthesis, those catalysing the uptake of nutrients, the ATPase responsible for<br />
maintaining the proton motive force <strong>and</strong> possibly receptors of cellular signalling systems.<br />
Membrane lipids are mainly the phospholipids, phosphatidyinositol (4.122), phosphatidylserine<br />
(4.121), phosphatidylcholine (4.118) <strong>and</strong> phosphatidylethanolamine (4.120). In<br />
addition, smaller quantities of sterols are present, the most abundant being ergosterol.<br />
The precise architecture of the plasma membrane remains uncertain. The fluid mosaic<br />
model (Singer <strong>and</strong> Nicholson, 1972) describes membranes as being bilayers of<br />
phosplipids in which the hydrophobic groups are turned inwards to face each other.<br />
The hydrophilic moieties are turned outwards <strong>and</strong> are in contact with the aqueous<br />
environment in the periplasm <strong>and</strong> cytoplasm, respectively. Proteins <strong>and</strong> sterols are<br />
located within the phospholipid bilayer. Sterols have polar hydroxy groups <strong>and</strong> a<br />
hydrophobic skeleton. This allows them to orientate themselves in a perpendicular plane<br />
between the hydrophobic chains of the phospholipids.<br />
The plasma membrane forms the barrier between the cytoplasm <strong>and</strong> the external<br />
environment. It prevents free diffusion of solutes <strong>and</strong> provides a support in which specific<br />
carrier proteins catalyse the selective uptake <strong>and</strong> excretion of metabolites. It provides a<br />
framework whose structure allows the generation of proton <strong>and</strong> ion gradients necessary<br />
for the generation of energy that drives many uptake reactions. An essential facet of<br />
cellular function is the ability to detect <strong>and</strong> respond to external stimuli. Receptors of<br />
cellular signalling systems are conveniently situated within the plasma membrane. The<br />
membrane provides a site where enzymes involved in various cellular synthetic pathways<br />
can be located in a manner that favours spatial arrangement <strong>and</strong> function.<br />
11.6.4 The cytoplasm<br />
The cytoplasm is that portion of the cell enclosed by the plasmalemma <strong>and</strong> excluding<br />
other membrane bound organelles. It is an aqueous colloidal liquid containing a multitude<br />
of metabolites. The cytoplasm of yeast is particularly rich in RNA. It is acidic, typically<br />
about pH 5.2, although localized metabolic activity may produce micro-environments of<br />
greater or lesser acidity. The enzymes of several major metabolic pathways are located<br />
within the cytoplasm, for example, glycolysis <strong>and</strong> fatty acid synthesis. Enzymes such as<br />
those of glycolysis are described as being `soluble' because in cell-free extracts the<br />
enzymes can be found in the supernatant when all membranous material has been<br />
removed by centrifugation. In fact, many of these enzymes are not r<strong>and</strong>omly dispersed<br />
throughout the cytoplasm. Instead, they are arranged in spatial configurations that aid<br />
ordered activity, possibly in loose associations with intracellular membranes.<br />
The cytoplasm contains a number of inclusions. Glycogen accumulates under<br />
appropriate conditions <strong>and</strong> appears in the cytoplasm as small granules that can be stained<br />
purple with iodine. Lipid particles become visible in the cytoplasm during aerobic growth<br />
when there is a plentiful supply of carbon. The particles apparently contain a hydrophobic<br />
core of triacylglycerol <strong>and</strong> steryl esters surrounded by a membrane consisting of<br />
phospholipid <strong>and</strong> protein (Leber et al., 1994). Probably lipid granules are temporary<br />
storage structures from which sterols may be transported to growing membranes <strong>and</strong><br />
triacylglycerols withdrawn in times of need.<br />
Ribosomes are cytoplasmic organelles that contain high concentrations of RNA <strong>and</strong><br />
some protein. Their role is to assemble proteins from `activated' amino acids sequenced<br />
in response to the code present in molecules of messenger RNA. Ribosomes are found<br />
throughout the cytoplasm either borne freely or often associated with the outer<br />
membranes of mitochondria, the endoplasmic reticulum <strong>and</strong> the outer nuclear envelope.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Commonly, several ribosomes are associated together joined by a str<strong>and</strong> of messenger<br />
RNA in structures termed polysomes.<br />
11.6.5 Vacuoles <strong>and</strong> intracellular membrane systems<br />
Yeasts, like other eukaryotes, contain extensive <strong>and</strong> dynamic internal membranous<br />
systems. These provide a method for partitioning metabolic pathways <strong>and</strong> pools of<br />
metabolites. They are also involved in the transport of metabolites both within the cell<br />
<strong>and</strong> to <strong>and</strong> from the plasma membrane. The most visible intracellular membranous<br />
system is the vacuolar system. Vacuoles are bodies bound by a membrane, the tonoplast.<br />
Their size <strong>and</strong> number fluctuates with physiological condition <strong>and</strong> stage in the cell cycle.<br />
When cells are growing in a balanced medium, as is the case during active primary<br />
fermentation, they may not be visible. Extensive vacuolation is associated with stress in<br />
yeast, especially starvation. Commonly large vacuoles become apparent in late<br />
fermentation or in stored pitching yeast.<br />
Vacuoles serve as temporary metabolite stores <strong>and</strong> provide the cell with a mechanism<br />
for controlling the concentration of metabolites in other cellular compartments. They are<br />
the site for the catabolism of macromolecules such as proteins. Vacuoles contain several<br />
proteinases, hence high concentrations of amino acids, especially basic types are also<br />
present. Thus, vacuoles function as a repository for nitrogen-containing metabolites.<br />
Their role as sites for proteolysis is consistent with the observation that they are most<br />
visible during starvation. The tonoplast, or vacuolar membrane, contains several amino<br />
acid transporters <strong>and</strong> is the site of a membrane-bound proton translocating ATP-ase. The<br />
latter is reportedly responsible for vacuolar acidification, which is an essential part of<br />
protein sorting (Klionsky et al., 1992).<br />
Vacuoles store quantities of inorganic phosphate, in the form of linear polymers of<br />
polyphosphate linked by high-energy phospho-anhydride bonds. Reportedly, polyphosphate<br />
is associated with S-adenosyl-L-methionine in vacuolar structures, termed volutin<br />
granules. These may be involved in the sequestration of basic amino acids (Schwenke,<br />
1991).<br />
The Golgi body comprises a dynamic series of stacked membranes <strong>and</strong> vesicles. It is<br />
part of the yeast secretory system <strong>and</strong> forms a link between the endoplasmic reticulum,<br />
the tonoplast <strong>and</strong> the plasma membrane. The endoplasmic reticulum consists of a<br />
branching network of membrane bound tubules. The Golgi body <strong>and</strong> the lumen of the<br />
endoplasmic reticulum are sites where proteins are sorted, modified <strong>and</strong> possibly<br />
assembled into complexes whilst being directed towards a chosen destination. Proteins,<br />
which are synthesized by ribosomes, may be transported across the membrane <strong>and</strong> into<br />
the endoplasmic reticulum. From here, they are directed towards the Golgi body via<br />
vesicles. In the Golgi body proteins are sorted <strong>and</strong> directed towards the sites where they<br />
are required. This may be within other intracellular organelles although most are sent to<br />
growing membranes. Many proteins, as evidenced by the numbers of visible vesicles, are<br />
sent to the area of membrane around the growing bud tip. Incorrectly folded proteins can<br />
be encapsulated in vesicles <strong>and</strong> returned to the endoplasmic reticulum. There they are<br />
repaired or degraded.<br />
Some proteins, encapsulated in vesicles, are transported across the plasma membrane <strong>and</strong><br />
secreted into the periplasm in a process termed exocytosis. A reverse process, termed<br />
endocytosis in which proteins are imported into the cell also occurs. The cytoplasm also<br />
contains a system of microtubules <strong>and</strong> microfilaments. These constitute the cytoskeleton <strong>and</strong><br />
are involved in the spatial organization of the cell especially during meiosis <strong>and</strong> mitosis.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Two cytoplasmic membrane-bound bodies, peroxisomes <strong>and</strong> glyoxysomes are<br />
associated with particular physiological states. Peroxisomes contain catalase <strong>and</strong><br />
oxidases required for the metabolism of specific carbon sources such as hydrocarbons<br />
<strong>and</strong> methanol. In glucose-grown cells of many yeast strains, including Saccharomyces<br />
spp, peroxisomes are barely evident. Transfer to amedium containing acarbon source<br />
suchasmethanolresultsintherapidbiogenesisofperoxisomesinthosestrainscapableof<br />
utilizing such substrates. Glyoxysomes are similar to peroxisomes but are rich in the<br />
enzymes of the glyoxylate cycle. Neither peroxisomes nor glyoxysomes are of relevance<br />
to brewing yeast under the conditions experienced during fermentation.<br />
11.6.6 Mitochondria<br />
Mitochondria are cytoplasmic organelles whose appearance <strong>and</strong> structure is much<br />
influenced by physiological state (Visser et al., 1995). Their principal role is energy<br />
generation via oxidative phosphorylation. They are most evident in cells growing<br />
aerobically under derepressed conditions (see Chapter 12). During oxidative derepressed<br />
growth, the mitochondrial volume (chondriome) accounts for some 12% of the total cell<br />
volume in S. cerevisiae. During anaerobic growth or under repressing conditions the<br />
chondriome is much reduced. Yeasts typically have a single or small number of large<br />
multi-branched mitochondria. Mitochondria have two membranes separated by an<br />
intermembrane space. The internal membrane has projections (cristae) that project into<br />
the internal matrix. Mitochondria contain a self-replicating genome, which codes for<br />
around 5% of all mitochondrial proteins. The remaining mitochondrial proteins are<br />
encoded by the nuclear genome <strong>and</strong> therefore biogenesis of these organelles requires coordinated<br />
expression of both sets of genes.<br />
Mitochondria are the site of the electron transport chain <strong>and</strong> oxidative phosphorylation<br />
resulting in the generation of ATP. In addition, there are enzyme systems for transporting<br />
many metabolites both into <strong>and</strong> out of the mitochondria, not least ATP. Several other<br />
enzyme systems unrelated to energy transduction are also located within this organelle<br />
including those of the oxidative tricarboxylic acid cycle, pyruvate dehydrogenase<br />
complex, several amino acid biosynthetic enzymes, the Mn-linked superoxide dismutase<br />
<strong>and</strong> possibly some of the sterol biosynthetic pathway. Under the repressing <strong>and</strong> anaerobic<br />
conditions of brewery fermentations, mitochondria do not contain the enzymes systems<br />
associated with oxidative metabolism. The genes coding for the proteins of the electron<br />
transport chain <strong>and</strong> several enzymes of the tricarboxylic acid are not expressed. The<br />
organelles remain in a partially undifferentiated state, termed promitochondria. These are<br />
small, difficult to visualize <strong>and</strong> internally few or no cristae are present. The chondriome<br />
is reduced to approximately 3±4% of the total cell volume (Stevens, 1977).<br />
Mitochondria do not have an energy-generating role during fermentation. Nevertheless,<br />
their presence is essential for normal fermentation behaviour because of the other<br />
metabolic functions they perform. O'Connor-Cox et al. (1996) have reviewed the<br />
functions of mitochondria, considered essential for normal fermentation performance.<br />
These include expression of flocculation, amino acid <strong>and</strong> diacetyl metabolism, sterol<br />
biosynthesis <strong>and</strong> physiological adaptation to stress.<br />
11.6.7 The nucleus<br />
The nucleus contains the greater part of the genetic material of the cell. It is roughly<br />
spherical, 1±2 micron ( m) in diameter <strong>and</strong> enclosed by a double membrane. The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
membrane is not continuous but contains several pores. The organelle is visible, with<br />
some difficulty, under the light microscope. In stationary phase cells it is often closely<br />
associated with vacuoles. The internal structure <strong>and</strong> appearance of the nucleus changes<br />
throughout the cell cycle. During interphase, acrescent shaped nucleolus is visible,<br />
associated with the nuclear membrane. The nucleolus is the site of rRNA transcription,<br />
some of the initial stages of mRNA processing <strong>and</strong> the assembly of ribosomal sub-units.<br />
The nuclear DNA of ahaploid cell is distributed between 16 linear chromosomes. The<br />
smallest chromosome (I) is 230kb in length, the largest (IV) 1532kb. The nuclei of most<br />
strains contain up to 100 copies of a2 mplasmid. This is acircular molecule of DNA<br />
<strong>and</strong> accounts for approximately 1% of the total nuclear DNA.<br />
Chromosomal DNA <strong>and</strong> proteins, both histones <strong>and</strong> non-histones are arranged<br />
together to form acomplex termed chromatin, so named because of its property of<br />
staining with basic dyes. In chromatin, the double helical DNA molecule is bound to a<br />
core of histone. Individual building blocks of DNA <strong>and</strong> histone are termed nucleosomes.<br />
Nucleosome units are coiled <strong>and</strong> condensed in hierarchical levels to form asupercoiled<br />
macromolecule. The protein component allows selective transcription of the genes on the<br />
chromosome. In comparison with higher eukaryotes, the chromatin molecules are loosely<br />
wound. This supports the view that the small genome is highly transcribed (Williamson,<br />
1991).<br />
The terminal regions of chromosomes are termed telomeres. These are involved in<br />
protecting the ends of chromosomes from degradation, maintaining the structural<br />
integrity of chromosomes, assisting in replication of DNA at the terminal region <strong>and</strong><br />
possibly in the attachment of chromosomes to the nuclear membrane during meiosis.<br />
Chromosomes also contain regions termed centromeres. These structures consist of DNA<br />
<strong>and</strong>protein<strong>and</strong>arethepartofthechromosomethatisattachedtothe spindlebodyduring<br />
mitotic division. Now the DNA component alone is called the centromere. The<br />
centromere is required to assemble alarge protein complex, termed the kinetochore. This<br />
structuremediatestheattachmentofchromosomestospindlemicrotubulesduringmitosis<br />
<strong>and</strong> ensures that each daughter cell receives acomplete set of chromosomes. Errors in<br />
chromosome segregation lead to alterations in chromosome copy number in daughter<br />
cells. This condition is termed aneuploidy. Interestingly, this is the common condition in<br />
the case of the genome of brewing yeast strains (Section 11.8.2).<br />
The spindle body acts as ascaffold, which mediates nuclear division <strong>and</strong> migration<br />
<strong>and</strong> segregation of daughter chromosomes during mitosis. It is adynamic structure that<br />
undergoes considerable changes during cellular budding. A complete spindle body<br />
consistsoftwospindlepolebodies<strong>and</strong>anumberofmicrotubules(Fig.11.6)Spindlepole<br />
bodies are electron dense plugs, approximately 0.15 m in diameter, which are located<br />
within the nuclear membrane. Short cytoplasmic (astral) microtubules project from a<br />
dense amorphous layer situated just above the spindle pole body. Two, or three, further<br />
types of microtubule are attached to a second amorphous dense area situated between the<br />
spindle pole body <strong>and</strong> the inner surface of the nucleus. Short discontinuous microtubules<br />
project a short distance into the nucleus. These are of two types. Firstly those that interact<br />
with the arms of chromosomes <strong>and</strong> secondly, those which interact with the kinetochore.<br />
Continuous microtubules form a direct link across the diameter of the nucleus between<br />
adjacent spindle pole bodies. All microtubules are approximately 20 nm in diameter.<br />
They all consist of repeating units of two proteins termed - <strong>and</strong> -tubulin.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Nuclear<br />
membrane<br />
Continuous<br />
microtubules<br />
11.7 Yeast cell cycle<br />
Spindle<br />
pole body<br />
Cytoplasmic<br />
microtubules<br />
Nuclear pore<br />
Chromatin<br />
Discontinuous<br />
microtubules<br />
Fig. 11.6 Diagrammatic representation of asection through the nucleus of ayeast cell in early G2<br />
phase.<br />
Proliferation of unicellular organisms involves coordination of the biochemical processes<br />
that together underpin growth of individual cells <strong>and</strong> those specific events that culminate<br />
in cellular multiplication. The combination of events that occur during the intervals<br />
between the separation of successive daughter cells is termed the cell cycle. It requires<br />
coordination of continuous events such as cellular growth with the discontinuous<br />
processes of DNA replication, mitosis <strong>and</strong> daughter cell excision. These processes in<br />
yeast have been subjected to intensive scrutiny.<br />
Progression through the cell cycle can be considered from three st<strong>and</strong>points. Firstly,<br />
themorphologicalchangesthatoccurasamothercellgivesbirthtoadaughter.Secondly,<br />
the biochemical events that underpin the process of cellular proliferation. Thirdly, the<br />
molecular mechanisms that regulate the coordinated processes of cellular growth <strong>and</strong><br />
multiplication. For convenience, the cell cycle is divided into anumber of phases (Fig.<br />
11.7). These are termed G1, which is the pre-synthetic gap phase; S, the synthetic phase<br />
during which DNA is replicated; G2, the post-synthetic gap phase, M, the mitotic phase<br />
<strong>and</strong> cytokinesis, the phase during which the daughter cell separates from the mother.<br />
During mitosis, the nucleus divides <strong>and</strong> duplicate pairs of chromosomes (chromatids) are<br />
sorted <strong>and</strong> segregated between the mother <strong>and</strong> daughter cell. In `classical' descriptions of<br />
mitosis, the process is delineated as aseries of distinct <strong>and</strong> named morphological states.<br />
In S. cerevisiae these steps are not clearly distinct.<br />
The gross morphology of mitotic cellular multiplication depends on the yeast type. In<br />
fission yeast, such as Schizosaccharomyces pombe, the mother cell divides to form two<br />
equal sized progeny. In the case of budding yeast such as S. cerevisiae, division is<br />
asymmetrical, the daughter cell being smaller than the mother. In some budding yeasts,<br />
the point where the bud forms is limited to specific points on the cell wall such as the<br />
poles. In others, including S. cerevisiae, buds may arise from any point on the cell wall.<br />
Nevertheless,budsdonotarisemorethanonceatthesamelocation.Thedeterminationof<br />
the precise site of bud initiation is aregulated process. Bud formation <strong>and</strong> separation is<br />
accompanied by nuclear division <strong>and</strong> segregation. This is adirectional process <strong>and</strong> thus,<br />
the cell must be polarized to ensure correct orientation of the dividing nucleus <strong>and</strong> the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CYTOKINESIS<br />
Separation of<br />
mother <strong>and</strong><br />
G1<br />
daughter Single spindle<br />
pole body<br />
Nuclear division<br />
ANAPHASE B<br />
Completion of spindle<br />
elongation in two phases<br />
(1 μm/min. <strong>and</strong> 0.3 μm/min.)<br />
START<br />
(Critical<br />
cell size<br />
achieved)<br />
developing bud. This is accomplished by the intermediary of the actin cytoskeleton <strong>and</strong><br />
the spindle body.<br />
Bud emergence occurs in response to localized degradation of the cell wall. The site of<br />
bud initiation is marked by the development of a ring structure, containing filaments<br />
termed septins (Harold, 1995). This process is accompanied by the accumulation of<br />
cytoplasmic vesicles in the area directly below the developing bud. These contain<br />
enzymes <strong>and</strong> precursors of cell wall synthesis. The vesicles are transported to the site<br />
using the actin microtubules of the cytoskeleton. The motive force for vesicle transport is<br />
provided by a multi-subunit protein termed dynein. Some of the sub-units of dynein<br />
provide an anchor attaching the vesicle `cargo' to the microtubule. Other subunits provide<br />
the motor force to move the vesicle from the positive to the negative end of the<br />
microtubule. The process is energized by the dissipation of ATP. The vesicles convey<br />
their contents to the site of bud growth by fusing with the plasma membrane in a process<br />
of exocytosis. The partial degradation of the wall allows cell turgor pressure to push out<br />
the developing bud.<br />
S<br />
Bud emerges<br />
DNA replication<br />
Spindle pole body<br />
replicated <strong>and</strong><br />
moves round<br />
nuclear envelope<br />
METAPHASE<br />
All chromatids<br />
attached <strong>and</strong><br />
oscillating round<br />
ANAPHASE A<br />
equatorial<br />
Segregation of chromatids at (metaphase) plate<br />
metaphase plate. Spindle pole<br />
bodies begin to move apart<br />
PRE-<br />
ANAPHASE<br />
Bi-polar spindle at<br />
2μm stage aligned<br />
with bud<br />
PRO-METAPHASE<br />
Spindle captures chromatids<br />
by binding centromeres to<br />
discontinuous microtubules<br />
via kinetochore<br />
Fig. 11.7 Phases in the cell cycle of S. cerevisiae (including brewing yeast strains).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Growth is restricted to the bud <strong>and</strong> there is no increase in the size of the mother cell<br />
during budding. Cell wall synthesis occurs outside the plasma membrane of the growing<br />
bud.Initially,cellwallgrowthisrestrictedtothetipofthedevelopingbud.Graduallycell<br />
wall synthesis spreads to the whole surface of the growing bud. At the end of mitosis,<br />
when the daughter nucleus <strong>and</strong> organelles such as mitochondria have divided <strong>and</strong> have<br />
migrated to the daughter cell, aseptum develops across the plane of the chitin ring. This<br />
process (cytokinesis) is completed when the daughter cell becomes detached from the<br />
mother.Theseptumonthemothercell iscomposedofchitin<strong>and</strong>constitutesthebudscar.<br />
The corresponding point on the cell wall of the daughter cell is deficient in chitin <strong>and</strong><br />
persists as the birth scar.<br />
Whilst bud emergence, growth <strong>and</strong> separation is occurring the nucleus <strong>and</strong> the<br />
enclosed chromosomes are duplicated <strong>and</strong> distributed between mother <strong>and</strong> daughter. In<br />
yeast, unlike mammalian cells, the nuclear membrane remains intact during the entire<br />
process. The morphological changesthatoccur duringreplicationofthe nucleus<strong>and</strong>their<br />
relation to bud formation <strong>and</strong> excision are illustrated in Fig. 11.7. Abrief description of<br />
the key events <strong>and</strong> associated nomenclature is also provided. The morphological changes<br />
associated with the cell cycle ofS. cerevisiae have been visualizedusing time-lapse highresolution<br />
digital enhanced differential interference contrast <strong>and</strong> multimode fluorescence<br />
microscopy (Shaw et al., 1998). These techniques provide direct visualization of the<br />
nucleus, the spindle <strong>and</strong> microtubules. Time lapse video footage of cells of S. cerevisiae<br />
progressing through the cell cycle may be found at http://www.molbiolcell.org. This<br />
elegant work shows how the discontinuous nuclear microtubules capture the duplicated<br />
chromosomes <strong>and</strong> segregate them at poles of the spindle body. The extension of the<br />
spindle body via elongation of the continuous microtubules is visualized. The manner in<br />
which cytoplasmic microtubules provide the motive power <strong>and</strong> orientation for movement<br />
of the dividing nucleus into the growing bud is demonstrated.<br />
Under the conditions used by these authors (Shaw et al., 1998) the average cell cycle<br />
time was 125 +/ 9 minutes. Of this, G1 <strong>and</strong> S together required 54 minutes to complete.<br />
Development of the short bipolar spindle (preanaphase) took a further 16 minutes.<br />
Capture <strong>and</strong> sorting of the daughter chromatids <strong>and</strong> elongation of the spindle to its fullest<br />
extent (anaphase A, B) required a further 30 minutes. The final division of the nucleus,<br />
migration of daughter nucleus into the bud <strong>and</strong> cytokinesis took 25 minutes. Molecular<br />
control of the cell cycle is predictably complex. Before the cell progresses into the S<br />
phase of DNA replication, it must achieve a critical size. The cell has mechanisms for<br />
sensing that an adequate supply of nutrients is available before it commits itself to<br />
replication. When these prerequisites are satisfied the cell cycle progresses through a<br />
critical checkpoint termed START. In S cerevisiae this is marked by the onset of bud<br />
emergence <strong>and</strong> DNA replication. If the cell is starved of an essential nutrient, it enters a<br />
resting state termed G0. The requirement to achieve a critical size explains why the<br />
average duration of the cell cycle is longer for daughter cells than their mothers. Thus, the<br />
former are smaller at birth <strong>and</strong> therefore require a longer G1 phase to achieve the critical<br />
size. A further checkpoint exists at the boundary between S <strong>and</strong> G2. At this point, there is<br />
a further critical cell size constraint, which must be satisfied before progression into<br />
mitosis occurs. In the case of S. cerevisiae, cells are usually sufficiently large at start to<br />
ensure that this second checkpoint is hidden.<br />
It can be shown to exist in the case of the fission yeast, Schizosaccharomyces pombe.<br />
This was demonstrated using the so-called wee mutants (Nurse, 1975). These cells have a<br />
gene, termed Wee 1, that is defective when the yeast is cultivated at high (restrictive)<br />
temperature. At the restrictive temperature, the cells undergo fission at a much smaller<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
size than the wild type. Thus, the normal product of the Wee 1gene restricts entry into<br />
mitosis. The function of this second checkpoint is to allow the cell to ensure that<br />
chromosome duplication <strong>and</strong> segregation has proceeded without error. If necessary, it<br />
provides an opportunity for corrective action. This is vital since errors at this stage could<br />
have fatal consequences. Progression through the cell cycle requires activation <strong>and</strong><br />
inactivation of enzyme systems via the intermediary of protein kinases <strong>and</strong> phosphatases.<br />
These regulatory cascade reactions are of the type described in Sections 12.2 <strong>and</strong> 12.5.8.<br />
Passage through the critical control points of the cell cycle is regulated by the controlled<br />
expression of genes that encode for catalytic proteins termed cyclins. At least 22 cyclins<br />
have been isolated. Cyclins interact with protein kinase cyclin-dependent kinases<br />
(CDKs).<br />
Several versions of these enzymes have been isolated from various cell types. In the<br />
case of S. cerevisiae, activation <strong>and</strong> deactivation of cdc28 (gene CDC28) regulates<br />
passage through the critical checkpoints of the cell cycle. Other cyclin-dependent kinases<br />
arealsoinvolved (Stark, 1999).Theprecisesequence ofeventsishighlycomplex<strong>and</strong>not<br />
fully characterized. Nevertheless, the general principles are that there is asensing system<br />
by which the cell monitors that the conditions are favourable for the cell cycle to<br />
progress. These include parameters such as nutrient availability, achievement of the<br />
critical cell size <strong>and</strong> confirmation that chromosome replication <strong>and</strong> segregation has been<br />
achieved. If the checks are satisfied, appropriate cyclins are synthesized <strong>and</strong> the cyclindependent<br />
kinase (cdc28) is activated. At START, for example, phosphorylation by the<br />
activated cdc28 kinase (S phase promoting factor) stimulates passage into the Sphase by<br />
activation of transcription. Mitosis promoting factor (cdc28 activated by B cyclins)<br />
controls passage from G2 to M.<br />
Individual cyclins are synthesized <strong>and</strong> degraded at various points in the cell cycle.<br />
Thus, G1 cyclins are associated with the passage from G1 to S. G2 cyclins (cyclin B)<br />
guide the cell through the transitions from Sto G2 <strong>and</strong> G2 into mitosis. Successive<br />
families of cyclins are responsible not only for promoting entry into the next stage of the<br />
cell cycle but also for degrading the cyclins responsible for driving the previous stage.<br />
Degradation is accomplished using the poly-ubiquination system associated with the<br />
yeast proteosome (Section 12.8).<br />
11.7.1 Yeast sexual cycle<br />
Many yeast genera are capable of sexual reproduction. S. cerevisiae is aperfect member<br />
of the Ascomycetae <strong>and</strong> is included in this group. Wild type strains of S. cerevisiae are<br />
usually diploid. Under appropriate conditions, these yeasts can be induced to undergo<br />
meiotic division <strong>and</strong> produce ascospores that are borne in afruiting body, an ascus.<br />
Industrial strains of S. cerevisiae, including brewing strains, are typically aneuploid/<br />
polyploid <strong>and</strong> do not normally have asexual cycle. From this st<strong>and</strong>point, the sexual stage<br />
of Saccharomyces yeasts is not relevant to brewing. Nevertheless, the sexual cycle of<br />
yeasthasbeenusedwidely, asamethodforexploringthegenome.Forthisreason,abrief<br />
descriptionoftheimportantfeaturesisincludedhere.Thekeyfeaturesoftheyeastsexual<br />
cycle are shown in Fig. 11.8.<br />
Haploid strains capable of sexual conjugation are of either of two mating types, termed<br />
MATa <strong>and</strong> MAT . Mating occurs in response to exposure of cells of opposite mating<br />
type to pheromones, termed a <strong>and</strong> factors, respectively. The pheromones are short<br />
peptides <strong>and</strong> they bind to specific receptor sites on the surface of cells of opposite mating<br />
type. Binding of the appropriate pheromone causes cells to arrest in the G1 phase of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
a/α<br />
Zygote<br />
a<br />
α<br />
A mating type<br />
haploid phase<br />
Diplophase<br />
a/α<br />
∝ mating type<br />
haploid phase<br />
HOMOTHALLIC<br />
spores fuse to<br />
give diploid<br />
Meiosis<br />
Asci with four<br />
ascospores<br />
HETEROTHALLIC<br />
Stable haploid phases<br />
formed providing cells<br />
of opposite mating<br />
type kept apart<br />
Fig. 11.8 Sexual life cycle of S. cerevisiae. Note that brewing strains of S. cerevisiae do not show<br />
a sexual cycle.<br />
cell cycle at START <strong>and</strong> thereby achieve cell cycle synchrony. Cells develop projections<br />
known as schmoos. At conjugation, the schmoos on adjacent cells become aligned, grow<br />
towards each other <strong>and</strong> eventually fusion occurs. This is mediated by expression of genes<br />
that encode for a specific group of surface agglutinins. Sexual conjugation is therefore a<br />
specific type of flocculation. During the fusion process, the plasma membranes of each<br />
cell become contiguous so the cytoplasm is shared. Nuclear fusion, or karyogamy to give<br />
a diploid cell follows cellular fusion. In a complete growth medium such diploid cells<br />
proliferate by budding <strong>and</strong> mitosis. This stage is known as diplophase. Under conditions<br />
of nutrient starvation or in the presence of a non-fermentable carbon source such as<br />
acetate or ethanol, but no nitrogen source, meiosis occurs <strong>and</strong> four haploid spores are<br />
formed. Spores can be induced to germinate by transfer to a rich growth medium.<br />
Two types of cell phenotype are possible depending on the possession or absence of a<br />
dominant allele termed HO. Heterothallic strains are HO <strong>and</strong>, providing the germinating<br />
ascospores are not brought into contact with a cell of opposite mating type, they will<br />
divide mitotically <strong>and</strong> produce stable haploid clones. Normally, ascospores are not<br />
liberated from the asci of S. cerevisiae, instead they fuse to produce diploid cells.<br />
Homothallic strains carry the dominant allele HO. Cells growing from spores, which<br />
bear the HO gene, are able to change mating type (Herskowitz et al., 1992). In budding<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
α<br />
a<br />
α<br />
a
cells,theswitchinmatingtypeoccursonlyinmothers<strong>and</strong>notdaughters.Oncethelatter<br />
havebecomemothers,theyalsoacquiretheability.Theswitchoccurswithanefficiency<br />
of approximately 60%. The result is that mating between siblings takes place <strong>and</strong><br />
therefore, the haploid phase in homothallic strains is transient. Mating type is<br />
determined by expression of genes at the MAT locus on chromosome III. In addition,<br />
chromosome III has two `silent' loci, or cassettes, which are located to the right <strong>and</strong> left<br />
of the centromere. These carry mating information for both MATa (HMR, right h<strong>and</strong> of<br />
centromere) <strong>and</strong> MAT (HML, left h<strong>and</strong> of centromere). As the terminology suggests,<br />
genes at the silent loci are not normally expressed. However, in some cases the genes at<br />
the MAT locus are exchanged with those from the silent HML or or HMR loci. These<br />
silent genes may then be expressed. Depending on the original MAT locus (either<br />
MATa or MAT ) <strong>and</strong> the substitute, mating type can be changed. In this set of<br />
circumstances the cells of opposite mating type will rapidly fuse <strong>and</strong> regenerate diploid<br />
cells.<br />
The mating process is induced by asignal transduction pathway involving aprotein<br />
kinase cascade termed MAPK (Stark, 1999). The mating pheromones bind to cell surface<br />
receptors <strong>and</strong> in so doing activate atetrameric protein (G protein) whose sub-units<br />
dissociate in response to exchange of GDP for GTP. Sub-units of Gprotein in turn<br />
activate the formation of a second multi-subunit complex, which triggers the MAP<br />
protein kinase cascade. In turn this interacts with cyclins with aresultant arrest in the cell<br />
cycle at START. Other MAP kinase target sites mediate the reactions leading to schmoo<br />
formation <strong>and</strong> fusion.<br />
Intheprocessofmeiosis,alsoknownasreductivedivision,thediploidnucleusdivides<br />
twice to form four haploid nuclei. In the initial fusion, two haploid nuclei fuse to form a<br />
single diploid zygote containing two sets of chromosomes. In sporulation medium, the<br />
two sets of chromosomes come together as homologous pairs (the two corresponding<br />
chromosomes from the original haploid parents). The DNA is then replicated, during<br />
which there is the opportunity for crossing over of genes between pairs of chromosomes.<br />
In this way genetic material can be shared to produce daughter cells with characteristics<br />
inherited from either parent. Aprocess of separation then follows involving homologous<br />
pairs of chromosomes <strong>and</strong> then individual replicas.<br />
Eventually four sets of chromosomes are formed each enclosed in anucleus. The<br />
process of sporulation begins when spore coat materials are synthesized <strong>and</strong> deposited<br />
around the nucleus. The cell enclosing the spores develops into aspore-bearing body, the<br />
ascus. Spore formation is accompanied by the accumulation of the carbohydrate,<br />
trehalose (Dickinson, 1988). This material makes cells resistant to stresses such as<br />
desiccation by virtue of its ability to stabilize membranes (Section 12.5.7).<br />
11.8 Yeast genetics<br />
Genetics is the study of the relationships between the structure of the genotype <strong>and</strong><br />
phenotypic expression. Genetic analyses provide a means of exploring the evolutionary<br />
<strong>and</strong> taxonomic relationships between individual strains. An underst<strong>and</strong>ing of the make-up<br />
of the genotype is a prerequisite for phenotypic modification. Thus, with knowledge of<br />
the nature of the genotype opportunities may present themselves by which undesirable<br />
characteristics can be deleted <strong>and</strong> desirable characteristics acquired. The comparatively<br />
rapid cell cycle of yeast, its ease of cultivation <strong>and</strong> relatively compact genome has made<br />
these organisms a common choice for the study of eukaryotic genetics, consequently, the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
scientific literature is enormous. Nevertheless, the majority of these studies employ<br />
haploid laboratory strains of S. cerevisiae. As discussed below, industrial strains of S.<br />
cerevisiae have amore complex genetic make-up. Aglossary of some terms used in<br />
genetics is given in Table 11.7 on pages 397±398.<br />
11.8.1 Methods of genetic analysis<br />
Genetic analyses are commonly performed using mutants. Exposure of cells to ionizing<br />
radiation or certain chemical compounds (mutagens) induces damage to the genome.<br />
Providing the treatment is not lethal, it is possible to obtain mutants. In other words, cells<br />
with an altered genotype with one or more defective genes. Selective media or selective<br />
cultural conditions can be used to identify <strong>and</strong> isolate mutants that are defective in the<br />
area of interest.<br />
The traditional method for determining the location <strong>and</strong> number of genes within the<br />
genome is by mating <strong>and</strong> meiotic recombination. Yeast is particularly suitable for this<br />
approach since the four ascospores are the result of a single meiotic event. Tetrad analysis<br />
allows mapping of genes relative to their centromeres <strong>and</strong> the drawing up of linkage maps.<br />
The latter indicate the relative distances between genes located on the same chromosome.<br />
Tetrad analysis is laborious. It involves mating of selected haploid strains to produce a<br />
hybrid diploid. After the induction of meiosis by transfer to sporulation medium,<br />
individual haploid ascospores are isolated by micromanipulation following enzymic<br />
digestion of the ascus wall. After induction of germination, the phenotype of the resultant<br />
haploid yeast lines can be assessed. Many aspects of the phenotypes of isolates can be<br />
assessed using the technique of replica plating. Here isolates are plated out onto a<br />
complete growth medium <strong>and</strong> incubated to allow the formation of colonies. An imprint of<br />
the colonies is made by pressing a piece of sterile velvet onto the plate. This is then used to<br />
inoculate plates of media that are selective for chosen phenotypic attributes. Comparison<br />
of patterns of growth on master <strong>and</strong> replica plates allows the assessment of phenotype.<br />
Tetrad analysis has allowed more than 90% of the genetic map of S. cerevisiae to be<br />
established (Cox, 1995). Although more modern techniques have superseded tetrad<br />
analysis as the method of choice for gene mapping it is still used to determine that a<br />
mutation has resulted from alteration of a single locus. Analysis of tetrads relies on the<br />
ratios of individual phenotypes to determine the relationships of the genes under<br />
investigation. Thus, a hybrid that is heterozygous for two markers, AB <strong>and</strong> ab has the<br />
potential to produce three classes of tetrad. These are: AB, AB, ab, ab (PD or parental<br />
ditypes); aB, aB, Ab, Ab (DPD or non-parental ditypes) <strong>and</strong> AB, Ab, ab, aB (T or<br />
tetratype). If the genes are unlinked, a r<strong>and</strong>om assortment will yield a ratio of 1 : 1 : 4 for<br />
PD, NPD <strong>and</strong> T, respectively. In the case of linked genes, there is an excess of PD to NPD<br />
tetrads. Where two genes are on different chromosomes <strong>and</strong> are linked to their respective<br />
centromeres the proportion of T type tetrads is reduced.<br />
The distance between individual genes on chromosomes influences the chances of<br />
crossovers during meiosis. Therefore, the frequency of occurrence of the different classes<br />
of tetrads can be used to determine the distance between genes. The distance between<br />
genes is measured in centiMorgans (cM). A centiMorgan is the unit on a genetic map<br />
equal to the distance along a chromosome that gives a recombination frequency of 1%.<br />
For map distances up to 35 cM the following equation may be used:<br />
cM ˆ 100<br />
2<br />
T ‡ 6NPD<br />
PD ‡ NPD ‡ T<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
For greater distances, up to 75 cM, a correction has to be made <strong>and</strong> the following<br />
empirically derived calculation is used:<br />
cM …corr:† ˆ<br />
…80:7† …cM† …0:883† …cM†2<br />
83:3 cM<br />
The distance between a gene <strong>and</strong> its centromere can be determined by measuring the<br />
percentage of T tetrads using a marker gene that is known to be tightly linked to the<br />
centromere. In this instance, the following formula is used:<br />
cM ˆ 100<br />
2<br />
T<br />
PD ‡ NPD ‡ T<br />
The proportion of the genome that resides in the nuclear chromosomes is described as<br />
being Mendelian since recombination events proceed according to Mendelian rules of<br />
inheritance. Other elements of the genome such as that found in mitochondria can also be<br />
detected by tetrad analysis. For example, non-Mendelian inheritance is observed for any<br />
character that is coded for by non-chromosomal DNA.<br />
Identification of a gene associated with a particular mutation can be accomplished by<br />
genetic complementation studies. This involves producing a range of diploid crosses by<br />
mating a haploid bearing the uncharacterized mutation with a range of characterized<br />
haploid mutants with the same phenotype as the test strain. A control diploid is produced<br />
using the test mutant <strong>and</strong> a normal wild type strain. The normal wild type phenotype in<br />
the control heterozygote confirms that the unknown mutation is recessive. The<br />
heterozygous cross that produces the mutant phenotype must be double recessive <strong>and</strong><br />
therefore contains the same mutation as the unknown. All other heterozygotes lose the<br />
original mutant phenotype indicating that the uncharacterized mutation was not present in<br />
any of the control test strains.<br />
The development of recombinant DNA technology has revolutionized the ease by<br />
which the yeast genome may be analysed <strong>and</strong> manipulated. Very powerful <strong>and</strong> precise<br />
techniques are now available that allow detailed analysis of the genome. Individual genes<br />
can be targeted, modified <strong>and</strong> transferred between cells to facilitate the study of the<br />
relationships between genotype <strong>and</strong> phenotype. It is now relatively simple to introduce<br />
foreign DNA into a cell to effect desirable changes in the phenotype. Detailed <strong>and</strong> rapid<br />
genetic analyses provide convenient <strong>and</strong> precise methods for strain identification ± socalled<br />
genetic fingerprinting.<br />
Many <strong>and</strong> varied techniques have been developed. They share some common themes.<br />
Unlike classical genetic techniques, DNA is extracted <strong>and</strong> manipulated in isolation. In<br />
order to work with complex mixtures of DNA it is necessary to have methods of<br />
separation, enrichment <strong>and</strong> identification. In a DNA molecule, each gene is encoded by a<br />
unique sequence of bases. In single str<strong>and</strong>ed form, this sequence can be targeted<br />
providing a probe is available that contains the complementary sequence. Occasionally,<br />
the DNA sequence of interest may constitute a very small proportion of the whole.<br />
Methods are available, which allow specific fragments to be selected <strong>and</strong> duplicated in<br />
vitro. In recent years, much effort has been directed towards identifying the base<br />
sequences of entire genomes. The enormous magnitude of this task has necessitated the<br />
development of automatic sequencing equipment. The prize is the ability to identify any<br />
gene within the genome. The structure of the protein expressed by the gene can be<br />
manipulated. Artificial genes can be created in vitro.<br />
Electrophoretic characterization of mixtures of DNA is achieved by loading onto a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
permeable gel made from a material such as polyacrylamide or agarose. The gel is<br />
submersed in an electrolyte <strong>and</strong> an electric current is applied. DNA molecules are<br />
charged molecules <strong>and</strong> in the electrical field, they migrate between electrodes. The extent<br />
of migration is dependent on molecular size <strong>and</strong> charge. Several different electrophoretic<br />
techniques are used which allow separation of different ranges of DNA molecule. These<br />
may range from whole chromosomes to small DNA fragments a few bases in length.<br />
Whole chromosomes can be separated using pulse field electrophoresis. Chromosomes<br />
can be visualized by staining with ethidium bromide <strong>and</strong> viewing under ultraviolet light.<br />
This procedure is termed karyotyping. It is used as a method of strain identification <strong>and</strong> to<br />
probe chromosome stability. More typically, the extracted DNA is broken into fragments<br />
using enzymes termed restriction nucleases. The latter are a large family of bacterial<br />
enzymes that break DNA molecules at particular known recognition sequences. The fragments<br />
can be separated by gel electrophoresis. In the technique termed Southern blotting, the<br />
separated fragments of DNA are transferred to a nylon or nitrocellulose membrane. During<br />
the process, the double-str<strong>and</strong>ed DNA molecules are denatured to give single str<strong>and</strong>s on the<br />
nylon sheet. The presence of specific DNA sequences can be detected by treating the<br />
membranes with probes consisting of complementary str<strong>and</strong>s of DNA attached to a<br />
visualizing system. The latter may be a radioactive label or a dye such as a fluorochrome.<br />
This procedure can be used as a method of strain identification in the technique termed<br />
restriction fragment length polymorphism (RFLP). Providing the complementary DNA<br />
probe is available it can be used to detect the presence of any gene within the test<br />
organism. The same approach can be used to probe mixtures of RNA molecules. In this<br />
case, termed Northern blotting, mRNA molecules are subject to electrophoresis. Specific<br />
RNA sequences are detected using a probe, which is a single str<strong>and</strong> of DNA.<br />
In many cases, it is necessary to enrich extracted DNA to increase the yield of<br />
particular genes of interest. This process is termed amplification <strong>and</strong> it can be<br />
accomplished in several ways. It can be performed in vitro using DNA polymerase, the<br />
enzyme that is responsible for DNA replication in the cell. The technique is known as<br />
PCR, from polymerase chain reaction. The extracted target DNA is denatured by heating<br />
<strong>and</strong> each single str<strong>and</strong> attached to a suitable primer. The latter is a short nucleotide<br />
sequence that targets the DNA of interest, such that any that is attached to the primer is<br />
duplicated. The entire duplication process takes just a few minutes. Repeated cycling<br />
allows multiple copies of the genes of interest to be made. In a short time, this<br />
amplification process results in several million copies of the original DNA being<br />
manufactured. PCR can be used to amplify the DNA fragments formed in the restriction<br />
fragment length polymorphism (RFLP) approach to strain identification. This combination<br />
of methods, termed amplified fragment length polymorphism (AFLP) increases the<br />
sensitivity of the genetic fingerprint.<br />
Exogenous DNA fragments can be introduced into a cell <strong>and</strong> thereafter inherited <strong>and</strong><br />
expressed with the cell's own genome, a process termed transformation. DNA may be<br />
introduced directly by first converting the target cell into a sphaeroplast. This is<br />
accomplished by enzymic removal of the cell wall in the presence of an osmotic<br />
stabilizer. Incubation of sphaeroplasts <strong>and</strong> DNA on solid media containing osmotic<br />
stabilizers provide conditions under which transformation <strong>and</strong> re-growth of the cell wall<br />
can occur. Newer procedures obviate the need for removal of the cell wall. For example,<br />
cell walls can be made permeable to DNA by treatment with lithium acetate.<br />
Alternatively, cell membranes can be made temporarily permeable to DNA by<br />
application of an electrical field in the process of electroporation. Exogenous<br />
mitochondrial DNA can be introduced into the mitochondria of host cells by high-<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
velocity bombardment. In this case, the DNA is contained within a tungsten microprojectile,<br />
which is fired into the mitochondrion using compressed air.<br />
The most usual method of introducing exogenous DNA into a cell is via the<br />
intermediary of a vector. These are small pieces of circular DNA termed plasmids. The<br />
DNA fragment of interest is integrated into the plasmid DNA. They have several<br />
advantages as transforming agents compared to the use of pure DNA. Several different<br />
plasmids are available which can be used to accomplish several different functions. These<br />
include insertion, deletion, alteration <strong>and</strong> expression of genes. All have marker genes so<br />
those host cells containing the plasmid can be selected. Commonly, they bear genes<br />
conferring resistance to selected antibiotics. Plasmids are usually of the shuttle type,<br />
referring to the fact that apart from being able to maintain themselves in yeast cells they<br />
have an origin of replication that allows generation of a high copy number in an<br />
alternative host such as E. coli.<br />
The ability to grow the plasmid in a bacterium provides a convenient method of gene<br />
amplification. When introduced into a yeast cell some plasmids integrate into a<br />
chromosome whilst others are able to replicate autonomously. These differences are<br />
utilized in different gene manipulations. For example, integrative types might be used<br />
where a very stable transformation is required. Conversely, autonomously replicating<br />
types may be used where very high copy numbers of the transforming gene are needed.<br />
Plasmids can be used to perform a number of tasks. They afford the most precise way of<br />
identifying a mutant gene. Thus, mutant cells are transformed with plasmids bearing a<br />
library of fragments of the wild type genome. Any cells with the wild type phenotype<br />
must contain a functional copy of the gene on the plasmid. The latter can be recovered<br />
<strong>and</strong> analysed for the presence of the appropriate gene. Similar procedures can be used to<br />
introduce a selected gene into a cell or to disrupt a chosen gene within a cell.<br />
The sequence of bases in DNA molecules can be determined. In order to deal with<br />
whole genomes, the enormous number of bases involved requires that processes are<br />
automated. Robotic samplers <strong>and</strong> automatic analysers are now available. Typically, the<br />
shotgun sequencing approach has been used. In this method, the genome is cut into a<br />
large number of r<strong>and</strong>om fragments. The sequence of each fragment is determined. Using<br />
the fast <strong>and</strong> relatively inexpensive computational power now available, the sequence of<br />
the whole can be deduced from those of the fragments.<br />
The entire genome of S. cerevisiae has now been sequenced <strong>and</strong> the DNA<br />
corresponding to each identified chromosomal gene is available in the form of a library.<br />
Pieces of apparatus termed arrayers have been developed that transfer, or print, individual<br />
polynucleotide samples corresponding to each gene onto a membrane. Each sample dot is<br />
approximately 100±200 m in size <strong>and</strong> arranged on a membrane in the form of a grid. The<br />
loaded grids hold the entire chromosomal genome. They are termed variously as DNA<br />
microarrays, DNA chips or gene chips. The oligonucleotides on the grids can be probed by<br />
hybridization with other samples of DNA. The probes are attached to fluorescent dyes <strong>and</strong><br />
hybridization can be visualized by confocal microscopy. Using this system, the DNA<br />
probe mixture can be prepared by reverse transcriptase from the entire mRNA content of<br />
the strain under investigation. By testing on a DNA array containing the original DNA<br />
genome of the strain, the relative expression of the whole genome can be assessed.<br />
11.8.2 The yeast genome<br />
Haploid cells of S. cerevisiae contain 16 chromosomes, ranging in size from 200 to<br />
2200 kb. All have been sequenced <strong>and</strong> 1996 saw publication of the complete sequence of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the entire genome of astrain of S. cerevisiae (Goffeau et al., 1996). The magnitude of this<br />
taskisreflectedbythefactthatinthecaseofS.cerevisiae,sometwelvemillionnucleotide<br />
bases were sequenced. Comparison of this sequence with those obtained from other cells<br />
<strong>and</strong> with knowledge of gene structure allows identification of sequences that encode for<br />
specific proteins. Such sequences are termed open reading frames (ORFs). Some 6,217<br />
potential open reading frames have been identified in the yeast genome (Mewes et al.,<br />
1997).Ofthese6,217potentialproteins,approximatelyhalfhavebeenpositivelyidentified<br />
basedonknownsequences.Afurther20%haveaputativeidentification<strong>and</strong>theremaining<br />
30%,theso-called`orphangenes'havenoknownfunction.Thepositivelyidentifiedgenes<br />
have been classified based on cellular function (Table 11.5).<br />
An attempt to gain an underst<strong>and</strong>ing of the function of orphan genes is being<br />
addressed by novel techniques. These involve the construction of mutants in which a<br />
single orphan gene is deleted. The phenotype is then examined to determine the<br />
metabolic consequences of the deletion. This approach has been termed reverse genetic<br />
analysis (Oliver, 1997). It is anticipated that this will culminate in the development of<br />
automatic analysis of the proteome in amanner analogous to the DNA array systems<br />
described in Section 11.8.1.<br />
Compared to many other cells, the yeast genome is very compact. Approximately 72%<br />
of chromosomal DNA codes for actual genes. The average size of yeast genes is 1.45 kb<br />
or 483 codons representing 40 to nearly 5,000 codons. Approximately 4% of yeast genes<br />
contain introns (non-coding regions). Genes are not evenly distributed throughout the<br />
chromosomal DNA, instead there are gene-rich clusters. In haploid strains, approximately<br />
half of the genes are duplicated. This has led to the suggestion that this species arose from<br />
the fusion of two ancestral diploid strains, each with eight chromosomes. The resultant<br />
tetraploid cell was reduced to a 16 chromosome diploid by deletion (Wolfe <strong>and</strong> Shields,<br />
1997). Yeast chromosomes contain additional DNA known as retrotransposons or Ty<br />
elements. Typically, laboratory haploid strains contain up to 30 copies of a number of<br />
different retrotransposons. These are related to retroviruses.<br />
Approximately 10% of the yeast genome is located in mitochondria. The<br />
mitochondrial genome encodes around 5% of mitochondrial proteins. The wild type<br />
mitochondrial phenotype is denoted + . Mutants termed ë lack all mitochondrial DNA.<br />
They are viable but lack many components of the electron transport chain <strong>and</strong> are<br />
therefore, respiratory deficient. Mutants that are ë produce small colonies on solid media<br />
<strong>and</strong> for this reason are termed petites. Petite mutation is relatively common in production<br />
brewing (Donnelly <strong>and</strong> Hurley, 1996). Strains of S. cerevisiae contain a variety of other<br />
Table 11.5 Yeast genome analysis of identified genes based on function (Mewes et al., 1997)<br />
Gene function Proportion of identified genome (%)<br />
Cellular organization <strong>and</strong> biogenesis 28<br />
Intracellular transport 5<br />
Transport facilitation 5<br />
Protein trafficking 7<br />
Protein synthesis 5<br />
Transcription 10<br />
Cell growth, division <strong>and</strong> DNA synthesis 14<br />
Energy transduction 3<br />
Metabolism 17<br />
Cell rescue 4<br />
Signal transduction 2<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
genetic elements. Most contain 2 mplasmids, which apparently function solely for their<br />
own replication since deletion has no observable effect on the yeast phenotype.<br />
Commonly, the cytoplasm of yeast contains viral nucleic acid termed dsRNA. Several<br />
types are found of which the M-type encodes for killer toxin.<br />
Much of the work, which has resulted in the current level of underst<strong>and</strong>ing of the yeast<br />
genome, has been performed on laboratory haploid strains. The genomes of industrial<br />
strains are very different. The most notable difference is the fact that brewing yeast strains<br />
are polyploid or aneuploid. Commonly, three or four sets of chromosomes are present<br />
(triploid,tetraploid).Oftenthesetsofchromosomesarenotpresentinmatchedsets,rather,<br />
one or more chromosomes is present as an extra or one less copy (aneuploid) (Hammond,<br />
1996). Polyploidy has many consequences. The sequenced haploid laboratory strain of S.<br />
cerevisiae contains a single copy of the genes encoding for the maltose fermenting<br />
enzymes, whereas brewing strains typically contain more than ten copies (Meaden, 1996).<br />
Polyploid strains of S. cerevisiae may have been selected for in industrial processes<br />
since they may have very stable phenotypes. Thus, the chance of asingle point mutation<br />
having an effect on the phenotype is reduced where multiple copies of the gene are<br />
present on other chromosomes. In addition, it is possible that multiple copies of some<br />
genes <strong>and</strong> concomitant increased expression might confer aselective advantage. For<br />
example, it has been claimed that multiple copies of maltose utilizing genes produces a<br />
phenotype where maltose utilization rates are higher than comparable haploid strains<br />
(Hammond, 1996).<br />
Noorganism has astable genotype. Mutation <strong>and</strong> selection provide the mechanism for<br />
evolutionary change. Although polyploidy reduces the chances of phenotypic change via<br />
mutation, it is not made impossible. <strong>Brewing</strong> yeast, in common with other fungi, exhibits<br />
chromosomal instability. This can be observed by karyotyping (Section 11.8.1). The<br />
changes show up as alterations in chromosome length (polymorphism) or even<br />
occasionally chromosome deletion. The changes in chromosome length are areflection<br />
of DNA rearrangement <strong>and</strong> distribution between individual chromosomes (Zolan, 1995).<br />
In brewing yeast, it has long been recognized that flocculation properties of individual<br />
strains are liable to abrupt shifts during repeated cycles of fermentation, cropping <strong>and</strong> repitching<br />
(Gillil<strong>and</strong>, 1971). Similar observations have been made with alager yeast strain<br />
used in production brewing. Genetic fingerprinting of the non-flocculent <strong>and</strong> flocculent<br />
variants showed differences (Wightman et al., 1996).<br />
11.9 Strain improvement<br />
Industrial strains of yeast, which have been in use for any length of time, will have been<br />
selected using empirical criteria. In brewing, for example, individual strains are used<br />
because they possess amix of desirable characteristics. These characteristics ensure that<br />
for particular combinations of wort, vessel type <strong>and</strong> yeast strain produce the desired<br />
fermentation performance <strong>and</strong> beer quality. Greater economic pressures of modern largescale<br />
brewing <strong>and</strong> the desire to develop new products have provided the impetus to<br />
improve strains. Improvement is taken to mean to introduce characters into existing<br />
strains such that they acquire an altered phenotype. Some of the potential goals of strain<br />
improvement are summarized in Table 11.6.<br />
Traditional genetic approaches for strain improvement involve mating between<br />
selected parents <strong>and</strong> selection of new hybrid offspring containing desirable traits.<br />
Alternatively, attempts can be made to select for mutants within a population that have<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.6 Goals for yeast strain improvement<br />
Aim Strategy<br />
Increased yield of ethanol Introduction of amylases to allow fermentation of wort<br />
dextrins<br />
Increased attenuation rate Increased concentration of glycolytic enzymes<br />
Very high gravity fermentation Increased ethanol <strong>and</strong> osmotolerance<br />
High temperature fermentations Increased thermotolerance without alteration in<br />
production of beer flavour metabolites<br />
Reduced risk of contamination by wild Incorporation of killer factor<br />
yeast<br />
Abolishment of diacetyl st<strong>and</strong> Increased flux through valine/isoleucine biosynthetic<br />
pathway<br />
Incorporation of -acetolactate decarboxylase<br />
Improved control of production of beer Manipulation of pathways for production of esters,<br />
flavour metabolites H 2S, SO 2<br />
Reduced ethanol yield Block ethanol production for low/zero alcohol beer<br />
production<br />
Altered flocculation Manipulation of Flo genes<br />
desirable characteristics. The frequency of occurrence of mutants can be increased by<br />
treatment withmutagens.<strong>Brewing</strong>yeaststrainsdonotlendthemselvestoeitherapproach<br />
because their polyploid nature makes them relatively resistant to mutation. In addition,<br />
they usually lack asexual cycle <strong>and</strong> sporulate at best poorly.<br />
It is possible to produce hybrids using rare mating (Spencer <strong>and</strong> Spencer, 1977;<br />
Young, 1981). This procedure is used to produce hybrids of respiratory deficient brewing<br />
strains <strong>and</strong> respiratory sufficient mutant strains. Hybrids are selected by ability to grow<br />
on defined media deficient in the nutrient requirement of the mutant but containing an<br />
oxidative carbon source. Using this technique, auxotrophic diploids with either aa or<br />
matingalleleshavebeenhybridizedwith brewingstrains.Thehybridswerecapableof<br />
sporulating. Providing one of the parents contains the kar allele it is possible to produce<br />
heteroplasmons. The kar allele prevents nuclear fusion <strong>and</strong>therefore these strains contain<br />
somatic elements of both parents with the nucleus of only one parent. This approach has<br />
been used to construct brewing strains containing killer factor.<br />
Protoplasts or sphaeroplasts are yeast cells which have had the cell wall removed by<br />
treatment with asuitable enzyme preparation. Providing the cells are suspended in an<br />
osmotically stabilized medium the sphaeroplasts do not lyse. In the presence of Ca 2+ ions<br />
<strong>and</strong> polyethylene glycol sphaeroplasts can be induced to fuse <strong>and</strong> form hybrids (Pesti <strong>and</strong><br />
Ferenczy, 1982). When sphaeroplasts are incubated on asuitable solid medium cell wall<br />
regeneration occurs <strong>and</strong> stable hybrids are obtained. Hybrids can be selected by ensuring<br />
that both parental types are auxotrophic for different nutrients. Only hybrids will grow on<br />
amedium that is deficient in both nutrients. The procedure can be used to produce stable<br />
tetraploid from diploid parents. It has been largely superseded by more amenable<br />
recombinant DNA techniques.<br />
Strain improvement using recombinant DNA technology has the enormous advantage<br />
of precision. It is possible to introduce or modify asingle gene within the target genome.<br />
Several methods may be used to introduce exogenous DNA into a recipient cell<br />
producing aphenotype with new <strong>and</strong> desirable attributes. The methods used are those<br />
outlined in Section 11.8.2. Usually the method of choice is one in which the exogenous<br />
DNA is introduced in the form of a plasmid vector of the type that forms a stable<br />
integrant with the host chromosomal DNA. These techniques have been applied<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.7 Glossary of terms used in genetics<br />
Term Definition<br />
Allelle One of a pair of genes located at the same point on homologous<br />
(syn. allelomorph) chromosomes<br />
Aneuploid Polyploid strains lacking whole copy numbers of all genes<br />
Ascospore Haploid spores produced via meiosis in Ascomycetous fungi<br />
Ascus Spore-bearing body in Ascomycetous fungi<br />
Auxotroph Mutant with specific nutrient requirement for growth<br />
Bases Constituent molecules of RNA <strong>and</strong> DNA whose order form the basis of<br />
genetic code. DNA contains adenine <strong>and</strong> guanine (purines); thymine <strong>and</strong><br />
cytosine (pyrimidines). In RNA uracil replaces thymine<br />
Biochip Array of 1000 or more oligonucleotides printed on a solid support. (syn.<br />
DNA chip, DNA microarray, gene chip, genome chip) used to identify genes<br />
present within a genome<br />
Cdc Cell division cycle mutants<br />
cDNA Complementary DNA produced by reverse transcription of mRNA<br />
Centromere Attachment site of chromosome to spindle body<br />
CHEF Electrophoresis technique ± contour clamped homogeneous electric field<br />
Chromosomes Structures in which duplex DNA molecules are located<br />
CLP Chromosome length polymorphism<br />
Codon Triplet of bases which together code for a single amino acid<br />
Differential Study of relative levels of all mRNA in cells in different physiological<br />
expression studies states (syn. transcriptiomics)<br />
Diploid Organism having two sets of chromsomes<br />
DNA Deoxyribonucleic acid ± macromolecule which forms the genetic code. It<br />
consists of a backbone of sugar <strong>and</strong> phosphate groups to which purine <strong>and</strong><br />
pyrimidine bases are attached. Two complementary molecules are attached<br />
together to form a double helix<br />
LTR Long terminal repeats ± long repeating sequences of bases flanking<br />
transposons in genome<br />
Meiosis Process of reductive division in which diploid nucleus divides with a single<br />
replication of DNA to give four haploid spores<br />
Mitosis Division of nucleus <strong>and</strong> replication of DNA during asexual cell division<br />
Monosomic Diploid cell lacking one of a pair of chromosomes<br />
Northern blot Electrophoretic technique using RNA instead of DNA<br />
ORF Open reading frame, sequences of DNA flanked by start <strong>and</strong> termination<br />
codons that code for a single protein<br />
Orphan genes DNA sequences identified as genes but not having recognized function<br />
PCR Polymerase chain reaction<br />
Plasmid Small, frequently circular sequence of extra-chromosomal DNA<br />
Probe Fragment of DNA of known sequence labelled with radioactive phosphorus<br />
or fluorochrome used for detection of complementary DNA in mixture<br />
Proteome Whole set of proteins produced by genome<br />
RFLP Restriction fragment length polymorphism<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 11.7 Continued<br />
Term Definition<br />
Rho Mutant lacking mitochondrial genome<br />
(syn.<br />
-<br />
)<br />
RNA Ribonucleic acid ± macromolecule consisting of a backbone of phosphate<br />
<strong>and</strong> sugar molecules to which purines <strong>and</strong> pyrimidines are attached. RNA<br />
exists in several functional forms:<br />
mRNA ± messenger RNA, single str<strong>and</strong>ed form which is synthesized during<br />
transcription of DNA. Provides template for amino acid synthesis<br />
rRNA ± ribosomal RNA, ribonucleic acid component of ribosomes, which<br />
mediate the process of translation of tRNA to form protein molecules<br />
tRNA ± transfer RNA, family of molecules which bind individual amino<br />
acids <strong>and</strong> transfer them to the ribosomes as directed by the codons on the<br />
mRNA molecule<br />
snRNA ± small nuclear RNA molecules that mediate post-translational<br />
processing of mRNA<br />
Sequence Order of nucleotide bases within molecules of DNA or RNA that collectively<br />
form the genetic code<br />
SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis. Gel<br />
electrophoresis separation technique for proteins. SDS causes dissociation<br />
of multimeric proteins<br />
Sporulation Spore formation. In S cerevisiae, ascospores are formed during meiotic<br />
division<br />
Southern blot Electrophoresis of DNA on gel followed by denaturation <strong>and</strong> transfer of<br />
single-str<strong>and</strong>ed DNA to membrane. Separated DNA fragments analyses by<br />
complementary DNA probes<br />
SSCP Single str<strong>and</strong>ed confirmational polymorphism ± separation technique similar<br />
to Southerm blotting differing in that it relies on the migrational<br />
characteristics of single-str<strong>and</strong>ed DNA<br />
TAFE Electrophoretic technique ± transverse alternating field electrophoresis<br />
Telomere Sequences of bases at end of eukaryotic chromosomes important for ensuring<br />
correct replication of terminal segments of DNA molecules<br />
Tetraploid Cell possessing four sets of chromosomes<br />
Transcription Synthesis of molecule of mRNA using single-str<strong>and</strong>ed DNA molecule as<br />
template<br />
Translation Synthesis of polypeptide in ribosome using mRNA as template<br />
Transposon Sequence of DNA that that is capable of movement within or between<br />
(transposable element) chromosomes (see Ty elements)<br />
Triploid Cell with three sets of chromosomes<br />
Trisomic Diploid cell containing three copies of one chromosome<br />
Ty elements DNA elements that are classified as long terminal repeat (LTR)-containing<br />
retrotransposons. These are retroviruses<br />
Western blot Technique similar to Southern <strong>and</strong> Northern blot for electrophoretic<br />
separation <strong>and</strong> detection of proteins<br />
Zygote Stage in sexual reproduction comprising the diploid cell formed from the<br />
fusion of two haploid cells of opposite mating type<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
successfully to brewing yeast strains. Recombinant yeast strains have been constructed<br />
which meet many of the needs outlined in Table 11.6 (Hammond, 1995).<br />
No yeast strains produced by recombinant genetic modification are used in production<br />
brewing even though one has been given approval for use by the United Kingdom<br />
regulatory authorities (Hammond, 1995). At present, at least within Europe, there is an<br />
extreme reluctance to accept genetically modified foods. Underst<strong>and</strong>ably, in this climate,<br />
commercial brewers are unwilling to risk the use of such yeast strains. Whether this<br />
situation will be transient or long lasting is unclear. It has had the result that development<br />
work in this area has been largely discontinued. Recombinant DNA technology continues<br />
to afford the most fruitful approach probing the secrets of the genome. Undoubtedly,<br />
within the short term considerable advances will continue to be made.<br />
11.10 References<br />
ANDERSON, R. G. (1995) 25th Cong. Eur. Brew. Conv., Brussels, 13.<br />
ARIAS, C. R., BURNS, J. K., FRIEDRICH, L. M., GOODRICH, R. M. <strong>and</strong> PARISH, M. E. (2002) Appl. Environ.<br />
Microbiol., 68, 1955.<br />
ARNOLD, W. N. (1991) Periplasmic space. InThe Yeasts, Vol. 4, 2nd edn, A. H. Rose <strong>and</strong> J. S. Harrison<br />
eds, pp. 279±295, Academic Press, London.<br />
ARVINDEKAR, A. U. <strong>and</strong> NARAYAN, B. P. (2002) Yeast, 19, 131.<br />
BACALLAO, R. <strong>and</strong> STELZER, E. H. K. (1989) Meth. Cell Biol., 31a, 454.<br />
BALDWIN, W. W. <strong>and</strong> KUBITSCHEK, H. E. (1984) J. Bacteriol., 158, 701.<br />
BARKER, M. G. <strong>and</strong> SMART, K. A. (1996) J. Amer. Soc. Brew. Chem., 54, 121.<br />
CAHILL, G., WALSH, P. K. <strong>and</strong> DONNELLY, D. (1996) J. Amer. Soc. Brew. Chem., 57, 72.<br />
CHAPMAN, A. C. (1931) J. Inst. <strong>Brewing</strong>, 37, 433.<br />
CORRAN, H. S. (1975) A History of <strong>Brewing</strong>. David & Charles, London.<br />
COX, B. S. (1995) `Genetic analysis in Saccharomyces cerevisiae <strong>and</strong> Schizosaccharomyces pombe'. In<br />
The Yeasts, Vol. 6, 2nd edn, A. E. Wheals, A. H. Rose <strong>and</strong> J. S. Harrison eds, pp. 7±67.<br />
CURTIS, N. S. (1971) Brewers' Guardian, September, 95.<br />
DE SOUZA PEREIRA, R., PARIZOTTO, N. A. <strong>and</strong> BARANAUSKAS, V. (1996) Appl. Biochem., Biotechnol., 59,<br />
135.<br />
DENGIS, P. B., NELISSEN, L. R. <strong>and</strong> ROUXHET, P. G. (1995) Appl. Environ. Microbiol., 103, 257.<br />
DICKINSON, J. R. (1988) Microbiol. Sci., 5, 121.<br />
DONNELLY, D. <strong>and</strong> HURLEY, J. (1996) Ferment, 9, 283.<br />
EDDY, A. A. (1958) `Aspects of the chemical composition of yeast'. In, The Chemistry <strong>and</strong> Biology of<br />
Yeasts, A. H. Cook, ed., pp. 157±249, Academic Press, London.<br />
FLEET, G.H. (1991) `Cell Walls'. In The Yeasts, Vol. 4, 2nd edn, A. H. Rose <strong>and</strong> J. S. Harrison eds, pp.<br />
199±277, Academic Press, London.<br />
FLORKIN, M. (1972) `A History of Biochemistry', in Comprehensive Biochemistry, 30, 129.<br />
FORGET, C. (1988) A Dictionary of Beer <strong>and</strong> <strong>Brewing</strong>, Brewers' Publications, Colorado, USA.<br />
GALITSKI, T., SALDANHA, A. I., STYLES, C. A., LANDER, E. S. <strong>and</strong> FINK, G. R. (1999) Science, 285, 251.<br />
GILLILAND, R. B. (1971) Brewers Guardian, Oct., 29.<br />
GOFFEAU, A., BARRELL, B. G. <strong>and</strong> BUSSEY, H. (1996) Science, 274, 546.<br />
GREUTER, W., BARRIE, F. R., BURDET, H. M., CHALONIER, W. G., DEMOULIN, V., HAWKSWORTH, D. I.,<br />
JOÈ RGENSEN, P. M., NICOLSON, D. H., SILVA, P. C., TREHANNE, P. <strong>and</strong> MCNEILL, J. (1996) International<br />
Code of Botanical Nomenclature, Koelz Scientific Books, KoÈnigstein, Germany.<br />
HAMMOND, J. R. M. (1995). Yeast, 11, 1613.<br />
HAMMOND, J. R. M. (1996). Yeast Genetics, In, <strong>Brewing</strong> Microbiology, 2nd edn, F. G. Priest <strong>and</strong> I.<br />
Campbell, eds, pp. 43±82, Chapman <strong>and</strong> Hall, London.<br />
HAROLD, F. M. (1995) Microbiology, 141, 2765.<br />
HAWKSWORTH, D. L. (1991) Mycological Res., 95, 641.<br />
HERSKOWITZ, I., RINE, J. <strong>and</strong> STRATHERN, J. N. (1992) `Mating-type determination <strong>and</strong> mating-type<br />
interconversion in S. cerevisiae'. In, The Molecular Biology of the Yeast Saccharomyces cerevisiae,<br />
Vol. 2 Metabolism <strong>and</strong> Biosynthesis, J. N. Strathern, E. W. Jones <strong>and</strong> J. R. Broach eds, pp. 583±656.<br />
Cold Spring Harbor Laboratory, USA.<br />
ISERENTANT, D. (1996) Cerevisia, 21, 30.<br />
JIN, Y. L. <strong>and</strong> SPEERS, A. (1998) Food Res. Internat., 31, 421.<br />
JOÈ RGENSEN, A. (1900) Micro-organisms <strong>and</strong> Fermentation, 3rd edn. Translated by A. K. Miller <strong>and</strong> A. E.<br />
Lennholm, Macmillan, London.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
KING, F. A. (1947) Beer Has a History, Hutchinson's Scientific <strong>and</strong> Technical Publications, London.<br />
KLIONSKY, D. J., NELSON, H. <strong>and</strong> NELSON, N. (1992) J. Biol. Chem., 267, 3416.<br />
KURTZMAN, C. P. <strong>and</strong> FELL, J. W. (1998) The Yeasts., A Taxonomic Study, 4th edn, (C. P. Kurzman <strong>and</strong> J.<br />
W. Fell eds., Elsevier, Amsterdam, Netherl<strong>and</strong>s.<br />
LAWRENCE, D. R., BOWEN, W. R., SHARPE, F. R. <strong>and</strong> VENTHAM, T. J. (1989) Proc. 22nd Cong. Eur. Brew.<br />
Conv., Zurich, 505.<br />
LEATHER, R. V., DALE, C. J. <strong>and</strong> MOROSON, B. T. (1997) J. Inst. <strong>Brewing</strong>., 103, 377.<br />
LEBER, R., ZINSER, E., ZELLNIG, G., PALTAUF, F. <strong>and</strong> DAUM, G. (1994) Yeast, 10, 1421.<br />
LLOYD, D. <strong>and</strong> CARTLEDGE, T. G. (1991). `Separation of yeast organelles', in The Yeasts, 2nd edn, vol 4,<br />
A. H. Rose <strong>and</strong> J. S. Harrison eds, pp. 121±174, Academic Press, London.<br />
LODDER, J. (1970) The Yeasts, a Taxonomic Study, 2nd edn, North Holl<strong>and</strong> Publishing Company,<br />
Amsterdam, Netherl<strong>and</strong>s.<br />
MEADEN, P. (1996) Ferment, 9, 213.<br />
MESTDAGH, M. M., ROUXET, P. G. <strong>and</strong> DUFOUR, J. P. (1990) Ferment, 3, 31.<br />
MEWES, H. W., ALBERMANN, K. <strong>and</strong> BARR, M. (1997) Nature, 387 (suppl), 7.<br />
MIKI, B. L. A., POON, N. H., JAMES, A. P. <strong>and</strong> SELIGNY, V.L. (1982) J. Bacteriol., 150, 878.<br />
MONTROCHER, R., VERNER, M.-C., BRIOLAY, J., GAUTIER, C. <strong>and</strong> MARMEISSE, R. (1998) Internat. J.<br />
Systematic Bacteriol., 48, 295.<br />
NARZISS, L. (1984) J. Inst. <strong>Brewing</strong>, 90, 351.<br />
NURSE, P. (1975) Nature, 256, 547.<br />
O'CONNOR-COX, E. S. C., LODOLO, E. J. <strong>and</strong> AXCELL, B. C. (1996) J. Inst. <strong>Brewing</strong>, 102, 19.<br />
OLIVER, S. G. (1997) Microbiology, 143, 1483.<br />
PESTI, M. <strong>and</strong> FERENCZY, L. (1982) J. Gen. Microbiol., 128, 123.<br />
PHAFF, H. J. <strong>and</strong> STARMER, W. T. (1987) `Yeasts associated with plants, insects <strong>and</strong> soil'. In The Yeasts 2nd<br />
edn, Vol. 1, A. H. Rose <strong>and</strong> J. S. Harrison eds, pp. 123±180, Academic Press, London.<br />
PISKUR, J., SMOLE, S., GROTH, C., PETERSEN, R. F. <strong>and</strong> PEDERSEN, M. B. (1998) Internat. J. Systematic<br />
Bacteriol., 48, 1015.<br />
QUAIN, D. E. (1986) J. Inst. <strong>Brewing</strong>., 92, 435.<br />
QUAIN, D. E. (1988) J. Inst. <strong>Brewing</strong>., 95, 315.<br />
REED, G. <strong>and</strong> NAGODARITHANA, T. W. (1991) Yeast Technology, Van Nostr<strong>and</strong>, USA.<br />
ROSE, A. H. <strong>and</strong> HARRISON, J. S. (1971) In The Yeasts, A. H. Rose <strong>and</strong> J. S. Harrison eds, 1st edn, pp. 1±2,<br />
Academic Press, London.<br />
SCHWENCKE, J. (1991) `Vacuoles, internal membranous systems <strong>and</strong> vesicles'. In The Yeasts, Vol. 4, 2nd<br />
edn, A. H. Rose <strong>and</strong> J. S. Harrison eds, pp. 347±432, Academic Press, London.<br />
SHAW, S. L., MADDOX, P., SKIBBENS, R. V., YEH, E., SALMON, E. D. <strong>and</strong> BLOOM, K. (1998) Molec. Biol. Cell,<br />
9, 1627.<br />
SINGER, S. J. <strong>and</strong> NICHOLSON, G. L. (1972). Science, 175, 720.<br />
SMITS, G. J., KAPTEYN, J. C., VAN DEN ENDE, H. <strong>and</strong> KLIS, F. M. (1999) Curr. Opinion in Microbiol., 2, 348.<br />
SPENCER, J. F. T. <strong>and</strong> SPENCER, D. M. (1977) J. Inst. <strong>Brewing</strong>, 83, 287.<br />
STARK, M. J. R. (1999). `Protein phosphorylation <strong>and</strong> dephosphorylation'. In, The Metabolism <strong>and</strong><br />
Molecular Physiology of Saccharomyces cerevisiae J. R. Richardson <strong>and</strong> M. Schweizer eds,<br />
pp. 209±275. Taylor & Francis, London.<br />
STEVENS, B. J. (1977) Biol. of the Cell, 28, 37.<br />
STRATFORD, M. (1992) Yeast, 8, 635.<br />
STRATFORD, M. (1994) Yeast, 10, 1741.<br />
STRATFORD, M. <strong>and</strong> ASSINDER, S. (1991) Yeast, 7, 559.<br />
TAKITA, M. A. <strong>and</strong> CASTILHO-VALAVACIUS, B. (1993) Yeast, 9, 589.<br />
TAMAI, Y., MOMMA, T., YOSHIMOTO, H. <strong>and</strong> KANEKO, Y. (1998) Yeast, 14, 923.<br />
TEUNISSEN, A. W. R. H. <strong>and</strong> STEENSMA, H. Y. (1995). Yeast, 11, 1001.<br />
VAUGHN-MARTINI, A. <strong>and</strong> MARTINI, A. (1993) Systematics <strong>and</strong> Appl. Microbiol., 16, 113.<br />
VISSER, W., VAN SPONSEN, E.A., NANNINGA, N., PRONK, J. T., KEUNEN, J. G. <strong>and</strong> VAN DIJKEN, J. P. (1995)<br />
Antonie van Leeuwenhoek., 67, 243.<br />
WIGHTMAN, P., QUAIN, D. E. <strong>and</strong> MEADEN, P. (1996) Lett. Appl. Microbiol., 22, 85.<br />
WILLIAMSON, D. H. (1991) `Nucleus, Chromosomes <strong>and</strong> Plasmids'. In The Yeasts, Vol. 4, 2nd edn, A. H.<br />
Rose <strong>and</strong> J. S. Harrison eds, pp. 433±488, Academic Press, London.<br />
WOLFE, K. <strong>and</strong> SHIELDS, D. (1997) Nature, 387, 708.<br />
YOUNG, T.W. (1981) J. Inst. Brew., 87, 292±295.<br />
YOUNG, T.W. (1987) `Killer Yeasts'. In The Yeasts, vol. 2, 2nd edn, A.H. Rose <strong>and</strong> J.S. Harrison eds, pp.<br />
131±164, Academic Press, London.<br />
ZOLAN, M. E. (1995) Microbiol. Rev., 59, 686.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
12<br />
Metabolism of wort by yeast<br />
12.1 Introduction<br />
The biochemical reactions that occur during fermentation represent the cumulative<br />
effects of the growth of yeast on wort. The disappearance of nutrients <strong>and</strong> the formation<br />
of ethanol, carbon dioxide <strong>and</strong> the other metabolites, which together contribute to beer,<br />
are all by-products of yeast growth. Wort is complex <strong>and</strong> not completely characterized.<br />
Similarly, the reactions which occur during fermentation are not fully characterized. The<br />
biochemistry of S. cerevisiae has been subject to intensive study. However, the majority<br />
of these studies involved laboratory strains growing under aerobic conditions, often on<br />
defined media. The literature describing the biochemistry of the metabolism of wort by<br />
brewing yeasts during brewing fermentation is much smaller. Here is provided an<br />
overview of the metabolism of S. cerevisiae. Where possible the discussion is limited to<br />
the metabolism of wort by brewing strains of S. cerevisiae. Some aspects of the<br />
metabolism of non-brewing strains of yeast are also included where they have relevance<br />
as potential spoilage organisms.<br />
<strong>Brewing</strong> yeast strains are heterotrophic, facultative anaerobes. Acomparatively wide<br />
spectrum of organic molecules can be oxidized to both generate energy in the form of<br />
ATP <strong>and</strong> simultaneously provide carbon skeletons for anabolic reactions. Depending on<br />
the availability of oxygen <strong>and</strong> the concentration <strong>and</strong> source of carbohydrate, metabolism<br />
may be fully aerobic <strong>and</strong> oxidative, or fermentative. Thus, brewing yeasts have a<br />
relatively versatile metabolism <strong>and</strong> are able to adapt to avariety of conditions <strong>and</strong><br />
furthermore, environmental triggers modulate their physiology. The biochemical events<br />
that occur during fermentation reflect the genotype of the yeast strain used <strong>and</strong> its<br />
phenotypic expression as influenced by the composition of the wort <strong>and</strong> the conditions<br />
established in the fermenting vessel. In order to obtain satisfactory fermentation<br />
performance <strong>and</strong> beer of adesired quality it is necessary to choose ayeast strain with a<br />
suitable genotype <strong>and</strong> manipulate the conditions to encourage appropriate metabolic<br />
behaviour.<br />
All brewing yeast strains have limited respiratory capacities <strong>and</strong> are subject to carbon<br />
catabolite repression (Gancedo, 1992; Section 12.5.8). Under the conditions encountered<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
during fermentation, this phenomenon ensures that yeast physiology is repressed at all<br />
times <strong>and</strong>, therefore, the major products of sugar catabolism are ethanol <strong>and</strong> carbon<br />
dioxide. The initial concentration <strong>and</strong> spectrum of fermentable carbohydrates control the<br />
concentration of ethanol synthesized during fermentation. The range of carbohydrates<br />
that the yeast is able to ferment <strong>and</strong> the maximum concentration of ethanol that it can<br />
tolerate are genetically determined. These factors are criteria used in the selection of<br />
yeast strains used in brewing.<br />
Besides ethanol <strong>and</strong> carbon dioxide, amultitude of other minor products of yeast<br />
metabolism are formed duringfermentation. Many ofthesecontribute tobeerflavour <strong>and</strong><br />
aroma.Theproductionofadesiredspectrumofflavour<strong>and</strong>aromacompoundsconstitutes<br />
another selection criterion for yeast strains, albeit with the caveat that fermentation<br />
conditions must be controlled properly to ensure that these minor metabolic by-products<br />
are synthesized in desired <strong>and</strong> consistent quantities.<br />
In the vast majority of breweries fermentation is abatch process. Usually, the pitching<br />
(inoculum) yeast is derived from a previous fermentation. This <strong>practice</strong> has many<br />
ramifications, some of which influence the metabolism of wort by yeast. Most notably, it<br />
provides a requirement to oxygenate wort. Molecular oxygen is necessary for the<br />
synthesis of sterols <strong>and</strong> unsaturated fatty acids. These are essential components of<br />
membranes <strong>and</strong> are aprerequisite for yeast growth under anaerobic conditions (Parks,<br />
1978; Section 12.6). Asingle dose of oxygen is supplied with the wort at the start of<br />
fermentation <strong>and</strong> this is used for the synthesis of sterol <strong>and</strong> unsaturated fatty acids.<br />
During the subsequent anaerobic phase of fermentation, the pre-formed pools of these<br />
metabolites, together with asmall quantity of lipid supplied with wort, are progressively<br />
diluted amongst daughter yeast cells. In the yeast crop obtained at the end of<br />
fermentation, sterol <strong>and</strong> unsaturated fatty acid levels are reduced to growth-limiting<br />
concentrations, hence, the need for oxygenation of wort in the next fermentation.<br />
At the end of fermentation, it is necessary to separate the yeast crop from the immature<br />
(`green') beer. This is facilitated by the design of the fermenting vessel <strong>and</strong> the flocculation<br />
characteristics of the yeast. Bottom-fermenting yeasts sediment to the base of the vessel at<br />
the end of fermentation, whereas top-fermenting types form acrop at the surface of the<br />
fermenting wort. Fermenter design must accommodate harvesting of each type of yeast.<br />
Typical high-gravity lager wort with aspecific gravity of 1.060 contains approximately<br />
150g/l fermentable sugar <strong>and</strong> 150mg/l free amino nitrogen. At the start of fermentation, the<br />
wort is oxygenated to achieve adissolved concentration within the range 15±25mg/l. It is<br />
pitched with yeast at arate of around 1gdry wt./l, equivalent to roughly 5gwet wt./l or<br />
12±15 10 6 cells/ml. Duringfermentation,theyeastconcentration increases aroundfivefold.<br />
Yeast growth is accompanied by the formation of roughly 45g/l ethanol <strong>and</strong> 42g/l<br />
carbon dioxide. The conversion of sugars to ethanol is about 85% of the theoretical. The<br />
shortfall represents the proportion ofwortsugars utilised for yeast biomassformation <strong>and</strong><br />
other metabolites. The yield of carbon dioxide is marginally less than that of ethanol<br />
since some of the former is fixed by yeast in carboxylation reactions. Growth of yeast on<br />
wort is an exothermic process <strong>and</strong> it is necessary to apply cooling during fermentation to<br />
dissipate heat. The changes in some key parameters during a high-gravity lager<br />
fermentation of this type are shown in Fig. 12.1.<br />
The <strong>practice</strong> of serial fermentation, cropping <strong>and</strong> re-pitching introduces a requirement<br />
for yeast storage in the intervals between fermentations. The duration of the storage phase<br />
<strong>and</strong> the conditions employed influence yeast physiological condition. In particular, the<br />
intracellular concentrations of storage carbohydrates (glycogen <strong>and</strong> trehalose) <strong>and</strong> sterols<br />
<strong>and</strong> other lipids can be influenced. Variations in the concentrations of these metabolites<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
(b)<br />
Dissolved oxygen<br />
concentration (mg l –1 )<br />
Yeast dry wt. (g l –1 )<br />
(c)<br />
5<br />
Yeast dry wt.<br />
1.06<br />
4<br />
1.05<br />
3<br />
2<br />
Suspended<br />
cell count<br />
1.04<br />
1.03<br />
1.02<br />
1<br />
PG<br />
1.01<br />
0<br />
0 50 100<br />
Time (h)<br />
150<br />
1.00<br />
200<br />
60<br />
40<br />
20<br />
Free amino nitrogen<br />
(mg l –1 )<br />
(d)<br />
Total higher alcohols<br />
(mg l –1 )<br />
Dissolved<br />
oxygen<br />
Temp.<br />
(Ethanol)<br />
0<br />
0 50 100<br />
Time (h)<br />
150<br />
0<br />
200<br />
280<br />
240<br />
200<br />
160<br />
120<br />
80<br />
40<br />
FAN<br />
Pyruvate<br />
20<br />
15<br />
10<br />
5<br />
Temperature (°C)<br />
PG<br />
60<br />
40<br />
20<br />
0<br />
60<br />
40<br />
20<br />
0<br />
Suspended cell count<br />
(×10 6 ml –1 )<br />
Fig. 12.1 Changes in present gravity (PG), yeast dry weight, suspended yeast cell count<br />
(a); ethanol concentration, temperature, dissolved oxygen concentration (b); free amino nitrogen<br />
concentration (FAN), pH, pyruvate concentration (c); total vicinal diketones (VDK), total higher<br />
alcohols, total esters (d) during the fermentation of a high-gravity (15 ëPlato) lager wort performed<br />
in a 1600 hl cyclindroconical fermenter. The fermentation was maintained at a temperature of 12 ëC.<br />
The process was terminated after approximately 150 h by cooling the green beer to 3 ëC.<br />
5.0<br />
4.8<br />
4.6<br />
4.4<br />
pH<br />
4.2<br />
0<br />
0 50 100<br />
Time (h)<br />
150<br />
4.0<br />
200<br />
80<br />
60<br />
H. alcohols<br />
Esters<br />
2.4<br />
2.0<br />
1.6<br />
40<br />
1.2<br />
0.8<br />
20<br />
VDK<br />
0.4<br />
0<br />
0 50 100 150<br />
0<br />
200<br />
Time (h)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
pH<br />
Total VDK (mg l –1 )<br />
120<br />
80<br />
40<br />
0<br />
20<br />
15<br />
10<br />
5<br />
0<br />
Pyruvate (mg l –1 )<br />
Total esters (mg l –1 )<br />
(Ethanol) (g l –1 )
in pitching yeast are acause of inconsistencies in the extent of growth during subsequent<br />
fermentation. Ipso facto, the conditions under which pitching yeast is stored should be<br />
such that the opportunity for physiological variability is minimized.<br />
The physiological state of pitching yeast is an important determinant of fermentation<br />
performance. Compared with other fermentations, pitching rates are comparatively high<br />
<strong>and</strong> growth extents are modest. The pitched yeast plays an active role in subsequent<br />
fermentation <strong>and</strong> aproportion may persist through to the crop <strong>and</strong> be subject to further<br />
rounds of storage <strong>and</strong> re-pitching. In amodern brewery, it is usual to limit the number of<br />
serial fermentations. When the permitted number of generations is reached, the yeast is<br />
discarded <strong>and</strong> anew pure culture is introduced into the brewery (Chapter 11). This<br />
<strong>practice</strong> necessitates the use of apure yeast culture plant <strong>and</strong> asystem for maintaining<br />
<strong>and</strong> cultivating reference cultures in the laboratory. Afeature of this approach is that a<br />
number of yeast `lines' of varying generational age will be in use at any given time.<br />
The influence of serial cropping <strong>and</strong> re-pitching on the consistency of fermentation<br />
performance <strong>and</strong> beer composition is unclear. Yeast cells undergo aprocess of ageing,<br />
senescence <strong>and</strong> death. The ageing process in yeast is associated with agradual disruption<br />
of many metabolic processes (Jazwinski, 1999). Depending on the type of fermenter,<br />
serial fermentation can select for asub-population enriched with elderly cells.<br />
12.2 Yeast metabolism ±an overview<br />
Metabolism is the sum of all the chemical processes occurring in a cell. The<br />
manifestations of metabolism are the disappearance of nutrients from the medium <strong>and</strong><br />
the appearance of by-products, of heat, the growth of individual cells <strong>and</strong> cell<br />
proliferation. These processes are accomplished by sequences of individual chemical<br />
reactions, which together form pathways. Each reaction is catalysed by functional<br />
proteins, termed enzymes. Metabolism is divided into two areas. Catabolism includes<br />
those pathways in which organic molecules are degraded with the liberation of energy.<br />
Anabolism is that part of metabolism in which the energy formed by catabolic pathways<br />
is utilized to fuel the synthetic reactions required for cellular growth <strong>and</strong> multiplication.<br />
Enzymes <strong>and</strong> other proteins are synthesized as aresult of the expression of individual<br />
genes (Chapter 11)<br />
Carbohydrates are the preferred sources of carbon <strong>and</strong> energy in yeast. The oxidation<br />
of carbohydrates liberates energy, furnishes reducing power <strong>and</strong> generates carbon<br />
intermediates. Some carbon intermediates, together with other non-carbohydrate<br />
nutrients, are utilized in anabolic metabolism to generate cellular biomass <strong>and</strong> byproducts.<br />
The energy is partly conserved in the form of the high-energy phosphate bonds<br />
of a group of metabolites, principally adensosine triphosphate (ATP, 4.53). Cleavage of<br />
these bonds liberates the stored energy <strong>and</strong> this is used to drive processes such as active<br />
transport <strong>and</strong> anabolic metabolism <strong>and</strong> to generate heat.<br />
Reducing power is transferred using the pyridine dinucleotide coenzymes,<br />
nicotinamide adenine dinucleotide (NAD + , 4.54) <strong>and</strong> to a lesser extent nicotinamide<br />
adenine dinucleotide phosphate (NADP + , 4.54). These compounds function as electron<br />
acceptors in reactions with oxidoreductase enzymes. In these reactions, two hydrogen<br />
atoms are removed from the substrate. One is released as a hydrogen ion <strong>and</strong> the second<br />
is transferred as a hydride ion to the nicotinamide portion of the coenzyme. The resultant<br />
reduced coenzyme (NAD(P)H) is then available to reduce a substrate <strong>and</strong> the oxidized<br />
form of the coenzyme is regenerated.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glucose<br />
Glycerol<br />
Triosephosphate<br />
NAD<br />
Lactate<br />
Pyruvate<br />
+ NADH + H +<br />
NADH + H +<br />
CH2 OH<br />
CH OH<br />
CH2 OH<br />
NADH<br />
Acetyl-CoA<br />
TCA cycle<br />
NAD + NADH + H +<br />
NADH + H +<br />
NAD +<br />
CH3 H<br />
C COOH<br />
CH3 CO COOH<br />
OH<br />
CH 3<br />
COO –<br />
Acetate<br />
NADH + H +<br />
NADH<br />
NAD +<br />
Acetaldehyde<br />
NAD +<br />
NADH<br />
Ethanol<br />
CH 3 CH2 OH<br />
Electron transport<br />
NADH + H +<br />
Amino acids<br />
α-Keto acids<br />
NADH + H +<br />
Higher alcohols<br />
α-Acetolactate<br />
Diacetyl<br />
NADH + H +<br />
Acetoin<br />
2,3-Butanediol<br />
An essential of cellular metabolism is the requirement to maintain redox balance. The<br />
oxidation reactions of carbohydrate dissimilation generate reduced pyridine nucleotides.<br />
The cell has afinite pool of pyridine nucleotides. In order to maintain activity of the<br />
glycolytic pathways the cell must ensure asupply of oxidized pyridine nucleotides. The<br />
ways in which this is accomplished do much to explain why certain products of<br />
metabolism accumulate (Fig. 12.2). Thus, in fully aerobic oxidative growth, NADH is<br />
reoxidized via the electron transport chain, which indirectly drives oxidative phosphorylation<br />
(Section 12.5.4). In this case the terminal electron acceptor is oxygen <strong>and</strong> water is<br />
formed. During fermentative growth, the oxidative pathways are inoperative <strong>and</strong> NAD +<br />
has to be regenerated by the reduction of acetaldehyde to ethanol (Section 12.5.5).<br />
Aproportion of the carbon metabolite flow through the carbohydrate dissimilatory<br />
pathways is diverted into anabolic biosynthetic pathways <strong>and</strong> so this material is not<br />
availabletotheterminalredox-balancingreactions<strong>and</strong>othermechanismscomeintoplay.<br />
The most often quoted is the formation of glycerol, via glycerol phosphate from<br />
dihydroxyacetone phosphate, although the formation of this metabolite is also considered<br />
aresponsetoosmoticstress(Section12.3.1).Duringbreweryfermentations,severalother<br />
options are also possible. For example, the terminal reductive step in the formation of<br />
higher alcohols from aldehydes <strong>and</strong> ketones derived from amino acids <strong>and</strong> the reduction<br />
of vicinal diketones. Most of these products of fermentation are important contributors to<br />
beer flavour.<br />
Metabolism is highly regulated. Control is exerted by the regulation of enzyme <strong>and</strong><br />
protein synthesis (gene level) <strong>and</strong> enzyme activity (phenotypic level). Spatial<br />
considerations are also important. Individual metabolic pathways are commonly localized<br />
in specific intracellular membrane-bound compartments. Transport of substrates to <strong>and</strong><br />
from these cellular compartments can control the activity of these pathways. Enzymes<br />
within a given cellular compartment probably have specific spatial relationships with one<br />
another <strong>and</strong> this too has regulatory significance.<br />
CH 3<br />
NAD +<br />
CHO<br />
CH3 C C CH3<br />
O O<br />
CH 3<br />
O2<br />
H2O<br />
H<br />
C C<br />
OH O<br />
CH3<br />
CH 3<br />
NAD +<br />
CH 3<br />
NAD +<br />
NADH + H +<br />
NAD +<br />
H H<br />
C C CH3<br />
OH OH<br />
C<br />
CH3 C COOH<br />
O OH<br />
Fig. 12.2 Some redox-balancing reactions utilizing the coenzyme, nicotinamide adenine<br />
dinucleotide (NAD + ). For clarity many reactions in the sequences of reactions have been omitted.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Control at the phenotypic level is achieved by varying enzyme concentration <strong>and</strong><br />
activity. Enzyme activity can be modulated by metabolites or proteins that exert<br />
stimulatory or inhibitory effects. These metabolites can be substrates or products of the<br />
enzyme in question. Alternatively, metabolites that are neither substrates nor products<br />
can alter enzyme activity by bringing about structural alterations, a process termed<br />
allosteric modulation.<br />
Some genes, termed constitutive, are always expressed <strong>and</strong> their protein products are<br />
present in the cell under all conditions. Other genes are expressed only under certain<br />
conditions. The control of the expression of these inducible genes may be simple. For<br />
example, the presence of agiven nutrient in the medium, e.g., maltose, commonly<br />
induces the expression of the genes which encode for the carriers (permeases) required<br />
for its transport into the cell <strong>and</strong> the enzymes involved in its subsequent utilization.<br />
Conversely, the accumulation of anutrient that is the product of an anabolic pathway<br />
often causes repression of the gene's coding for the proteins required for its synthesis.<br />
Othercontrolmechanismsinfluencebothgeneexpression<strong>and</strong>enzymeactivityinacoordinated<br />
fashion. These metabolic signal transduction systems have far-reaching effects<br />
on metabolism <strong>and</strong> are of fundamental importance in explaining why <strong>and</strong> how yeasts<br />
respond to different environmental stimuli (Thevelein, 1994). In these metabolic control<br />
systems one, or asmall number, of exogenous triggering chemical compounds in the<br />
medium interact with specific targets within yeast cells. This initial interaction sets in<br />
motion aseries of metabolic events in which several enzymes <strong>and</strong> transport permeases<br />
may be activated or inactivated. Simultaneously, the expression of several genes may be<br />
induced or repressed. Thus, exposure of yeast to asingle metabolite has the potential to<br />
alter several facets of its physiology. Conversely, a number of different exogenous<br />
triggers may activate a common pathway. Several signal transduction pathways are<br />
known tooperate inyeastcells <strong>and</strong>,nodoubt,morewillbeidentified inthe future. Often,<br />
there is overlap between signal transduction pathways. The regulation of individual<br />
pathways is organized in ahierarchical fashion such that activation <strong>and</strong> inhibition is<br />
ordered with respect to time. In this way, the cell organizes metabolic activity to suit any<br />
particular set of environmental conditions.<br />
Signal transduction pathways are commonly mediated by the phosphorylation or<br />
dephosphorylation of target proteins in response to the activities of specific protein<br />
kinases <strong>and</strong> phosphoprotein phosphatases (Stark, 1999). The attachment of aphosphate<br />
group to susceptible proteins leads to areversible structural modification in which the<br />
catalytic activity can be eliminated or enhanced. Often, the direction of flow through a<br />
metabolic pathway is regulated by pairs of enzymes whose activity is modulated by<br />
phosphorylation or dephosphorylation (Fig. 12.3). The importance of protein<br />
phosphorylation in the control of cellular metabolism is emphasized by the fact that<br />
some 120 protein kinases have been identified in S. cerevisiae, accounting for 2% of the<br />
genome.<br />
12.3 Yeast nutrition<br />
Some chemical components present in the wort or other medium surrounding yeast cells<br />
may be used as nutrients, some may be toxic or growth-suppressing, others may have no<br />
effect whatsoever. In some cases, the same component may be a nutrient at a low<br />
concentration or toxic at a higher concentration. Some substances are assimilated only<br />
under particular cultural conditions. Major classes of nutrients such as sources of carbon<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Protein kinase<br />
Degradation<br />
Protein kinase<br />
ATP<br />
ADP<br />
ADP<br />
ATP<br />
E1<br />
(inactive)<br />
E1P<br />
(active)<br />
E2P<br />
(inactive)<br />
E2<br />
(active)<br />
Phosphoprotein<br />
phosphatase<br />
Synthesis<br />
Phosphoprotein<br />
phosphatase<br />
Fig. 12.3 A representation of a system for the regulation of flow through a metabolic pathway, in<br />
the direction of biosynthesis or breakdown, using reversible phosphorylation <strong>and</strong> dephosphorylation.<br />
E 1 ˆ enzyme 1, E 1P ˆ Phosphorylated enzyme 1, E 2 ˆ enzyme 2, E 2P ˆ phosphorylated<br />
enzyme 2.<br />
<strong>and</strong> nitrogen are assimilated in an ordered fashion. Thus, where several sources of carbon<br />
<strong>and</strong> nitrogen are present yeast first utilizes those which are most readily assimilated.<br />
12.3.1 Water relations<br />
All yeasts require an aqueous medium for growth. The concentration of available water<br />
elicits different responses in individual strains. The available water, or water activity (aw),<br />
is an inverse function of the concentration of solutes present in the medium. Media with<br />
low water activities (high solute concentrations) are stressful <strong>and</strong> only a limited number<br />
of yeast strains can tolerate these conditions. It is difficult to distinguish between the<br />
effects due solely to water availability <strong>and</strong> those in which the nature <strong>and</strong> concentration of<br />
the solutes are the predominant influences. A solute may exert a specific biochemical<br />
influence on the physiology of the yeast. Neutral solutes exert a general osmotic stress,<br />
whereas charged species introduce an additional electrochemical factor. The way in<br />
which the stress is applied is also important. A very rapid change in the water activity of<br />
the medium has the greatest potential for damage.<br />
The multiplicity of mechanisms by which water activity may influence yeast<br />
physiology has resulted in a complicated <strong>and</strong> confusing system of nomenclature. Strains<br />
which can grow in media with low water activities <strong>and</strong> where this parameter alone is<br />
considered to be the major influence, are said to be xerotolerant. The term osmotolerance<br />
is used where the osmotic effect due to dissolved solutes is considered most important.<br />
Both of these terms describe strains that can tolerate extreme conditions but prefer a<br />
medium with a more moderate water activity. The descriptor, osmoduric is also used for<br />
such strains. Rarely, some strains grow best in media with a low water activity. These are<br />
classified as being osmophilic or xerophilic.<br />
Yeast strains, which can tolerate low water activity conditions, include Debaromyces<br />
hansenii, Hansenula anomola, Pichia ohmeri, Schizosaccharomyces pombe, Torulopsis<br />
c<strong>and</strong>ida, Zygosaccharomyces rouxii <strong>and</strong> Zygosaccharomyces bisporus (Rose <strong>and</strong><br />
Harrison, 1995). In terms of brewing, all of these are `wild' yeast strains.<br />
Some marine strains of Metschnikowia have been described which are reportedly<br />
obligate osmophiles (Phaff et al., 1978). Z. rouxii is of importance in brewing since it is<br />
the most common cause of microbial spoilage of concentrated sugar syrups. S. cerevisiae<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Pi<br />
Pi
is not considered to be xerotolerant. This `deficiency' places an upper limit on the<br />
concentration of wort that can be used. However, S. cerevisiae can withst<strong>and</strong><br />
dehydration. Yeast used for baking <strong>and</strong> wine making is routinely supplied in adried<br />
form <strong>and</strong> this technique has now been extended to brewing yeast (Fels et al., 1999). The<br />
dehydration process results in morphological changes. The cells take on ashrunken<br />
appearance <strong>and</strong> the plasma membranes develop deep invaginations (Rapoport et al.,<br />
1995). These changes are reversed during rehydration. Survival rates are low <strong>and</strong> for<br />
successful drying the yeast musthave been cultivatedunder aerobic fed-batch conditions.<br />
These cultural conditions are associated with elevated intracellular concentrations of<br />
trehalose <strong>and</strong> sterols, both of which probably protect membrane integrity during<br />
dehydration (Sections 12.5.7 <strong>and</strong> 12.7.3).<br />
Yeast responds to changes in the water activity of the suspending medium by transient<br />
changes in cellular volume. This depends on acombination of the inherent elasticity of<br />
the cell wall <strong>and</strong> an uncontrolled adjustment in the volume of intracellular water, which<br />
occurs in response to alterations in external osmotic potential (Marechal et al., 1995).<br />
Following ahyperosmotic shock (suspension in amedium with an osmotic potential<br />
higher than intracellular contents), the cells contract then there is agradual recovery<br />
period during which the cellular volume increases <strong>and</strong> returns to the initial value.<br />
Osmotolerant strains are more resistant to uncontrolled water loss following<br />
hyperosmotic shock <strong>and</strong> better able to control cellular volume (Walker, 1998).<br />
Xerotolerant strains withst<strong>and</strong> the conditions of low water activity by altering<br />
intracellular conditions. This is accomplished by the synthesis of neutral polyols such as<br />
arabitol, mannitol, erythritol, sorbitol <strong>and</strong> especially glycerol. These are termed<br />
compatible solutes. The physiological adaptation to hyperosmotic shock is termed the<br />
osmostress response. Intracellular accumulation of compatible solutes reduces the<br />
difference in osmotic potential between the interior of the cell <strong>and</strong> the environment. The<br />
synthesis of compatible solutes allows the re-establishment of cell volume <strong>and</strong> aids the<br />
stabilization of enzymes <strong>and</strong> membrane proteins <strong>and</strong> phospholipids (Mager <strong>and</strong> Varela,<br />
1993). Xerotolerant strains are able to retain compatible solutes within the cell. Less<br />
xerotolerant types, including S. cerevisiae, lack this ability <strong>and</strong> some leakage occurs. S.<br />
cerevisiae also differs from more xerotolerant species in that following ahyperosmotic<br />
shock, the accumulation of glycerol is preceded by the formation of trehalose <strong>and</strong><br />
accumulation of potassium <strong>and</strong> sodium ions (Singh <strong>and</strong> Norton, 1991).<br />
The release or retention of glycerol, in response to changes in osmotic potential, is<br />
regulated.InZ,rouxii,aspecificactiveglyceroltransporterhasbeenidentified(VanZyl<br />
et al., 1990). In S. cerevisiae, the presence of aplasma membrane channel has been<br />
described through which the passage of glycerol is controlled in response to the osmotic<br />
potential of the medium (Luyten et al., 1995). Glycerol accumulates in beer during<br />
fermentation, typically 1±2g/l being formed. The osmotic stress response in S.<br />
cerevisiae is mediated by aspecific sensing pathway termed the HOG (high osmolarity<br />
glycerol) signal transduction pathway. Two separate osmosensors located in the plasma<br />
membrane activate a common MAP (mitogen-activated protein) kinase signal<br />
transduction cascade. In response to the osmotic shock, ametabolic signal is passed<br />
through the kinase cascade, which culminates in the transcription of genes that encode<br />
enzymes including those responsible for glycerol synthesis (Maeda et al., 1994). The<br />
HOG signal transduction pathway activates genes, which contain an element termed<br />
STRE in their promoter regions. The STRE element is present in many genes associated<br />
with responses to stresses other than osmotic shock. This general stress response is<br />
discussed in Section 12.9.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
12.3.2 Sources of carbon<br />
Yeasts, as a group, can assimilate a comparatively wide range of organic carbon<br />
compounds. An assessment of the range of such compounds that are utilized by individual<br />
strains is used as adiagnostic criterion in yeast taxonomy (Barnett et al., 1990; Chapter<br />
11). The most commonly used carbon sources are carbohydrates, including mono-, di- <strong>and</strong><br />
trisaccharides, higher dextrins <strong>and</strong> starches. Some species can utilize pentoses, although<br />
not brewing yeast strains. Growth may be oxidative <strong>and</strong>/or fermentative depending on the<br />
strain <strong>and</strong> the cultural conditions. Less usual carbon sources are used by certain species<br />
such as methanol <strong>and</strong> both aliphatic <strong>and</strong> aromatic hydrocarbons.<br />
Strains of S. cerevisiae utilize a limited repertoire of carbon sources for growth (Table<br />
12.1). Differences in the patterns of utilization are strain-specific. Ale strains lack the<br />
ability to utilize the disaccharide, melibiose. Lager strains can grow on melibiose because<br />
they have -D-galactosidase activity, which hydrolyzes it to galactose <strong>and</strong> glucose<br />
(Barnett, 1981). There are other differences between ale <strong>and</strong> lager strains. The latter<br />
utilize maltotriose more rapidly than ale strains (Stewart et al., 1995) <strong>and</strong> lager strains are<br />
more efficient at assimilating galactose. Lager strains utilize mixtures of galactose <strong>and</strong><br />
maltose simultaneously, whereas ale strains assimilate maltose preferentially (Crumplen<br />
et al., 1993). The yeast, S. cerevisiae var. diastaticus can utilize dextrins, albeit with the<br />
chemically unrelated formation of styrene <strong>and</strong> 4-vinyl guaiacol which impart a phenolic<br />
medicinal taint to beer (Ryder et al., 1978).<br />
Many metabolic intermediates accumulate in the growth medium. These include<br />
pyruvate (4.146), acetaldehyde <strong>and</strong> several organic acids such as citric (4.153), malic <strong>and</strong><br />
acetic. Some of these may be re-assimilated at a later stage in growth. The major products<br />
of fermentative growth are ethanol, glycerol, lactic acid <strong>and</strong> carbon dioxide. A proportion<br />
of the latter is fixed in a series of carboxylation reactions. Up to 5% of the carbon<br />
Table 12.1 Utilization of carbon sources by S. cerevisiae under aerobic <strong>and</strong> anaerobic conditions<br />
Carbon source Aerobic Anaerobic<br />
D-Glucose All strains All strains<br />
Cellobiose ± ±<br />
Ethanol Some strains ±<br />
D-Galactose Some strains ±<br />
D-Glucitol ± ±<br />
Glycerol Some strains ±<br />
Inulin ± ±<br />
DL-Lactate Some strains ±<br />
Lactose ± ±<br />
Maltose Some strains Some strains<br />
D-Mannitol Some strains ±<br />
Melezitose Some strains Some strains<br />
Melebiose Some strains Some strains<br />
Methanol ± ±<br />
Methyl- -D-glucopyranoside Some strains Some strains<br />
Raffinose Some strains Some strains<br />
L-Sorbose ± ±<br />
Starch Some strains Some strains<br />
Succinic acid Some strains ±<br />
Sucrose Some strains Some strains<br />
Trehalose Some strains Some strains<br />
D-Xylose ± ±<br />
Xylitol ± ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
equirement of yeast is provided by carbon dioxide fixation (Oura et al., 1980). In the<br />
presence of oxygen, some strains can undergo ametabolic adaptation, termed the diauxic<br />
shift, <strong>and</strong> utilize ethanol, lactic acid <strong>and</strong> glycerol for oxidative growth.<br />
12.3.3 Sources of nitrogen<br />
Yeasts cannot assimilate gaseous nitrogen, however, simple inorganic sources such as<br />
ammonium salts may be readily utilized. Adiverse range of organic sources of nitrogen<br />
can be assimilated (Soumalainen <strong>and</strong> Oura, 1971) including amino acids, peptides,<br />
amines, pyrimidines <strong>and</strong> purines. Many of these, for example, amines, are utilized as a<br />
source of nitrogen only in the presence of additional sources of carbon <strong>and</strong> energy. The<br />
ability, or inability, to use a specific organic nitrogen source can have taxonomic<br />
significance. Saccharomyces yeasts cannot utilize nitrate or nitrite but readily assimilate<br />
ammonium ions. In natural media, such as brewers' wort, ammonium ions, amino acids,<br />
peptides, purines <strong>and</strong> pyrimidines provide most of the nitrogen. These yeasts strains do<br />
not produce extracellular proteases <strong>and</strong> therefore, proteins are not utilized.<br />
12.3.4 Sources of minerals<br />
Sulphur may be assimilated from both inorganic <strong>and</strong> organic sources. The latter include<br />
the sulphur-containing amino acids, methionine (4.41) <strong>and</strong> cysteine (4.31). In addition,<br />
glutathione may be assimilated. The preferred inorganic source of sulphur is sulphate but<br />
sulphite <strong>and</strong> thiosulphate can also be utilized. S. cerevisiae can reduce elemental sulphur<br />
to thiol ions in the periplasm <strong>and</strong> thereafter these are assimilated (Rose, 1987). The<br />
requirement for phosphorus is satisfied by the assimilation of inorganic phosphate ions.<br />
Elemental mineral ions are essential co-factors for numerous enzyme activities. Some<br />
have structural roles <strong>and</strong> others are necessary components of transport systems where<br />
they fulfil acharge-balancing role. The concentrations required for growth are small,<br />
typically less than 10 M(Jones <strong>and</strong> Greenfield, 1984). Essential mineral ions include<br />
B + , Ca 2+ , Co 2+ , Cu 2+ , Fe 3+ , K + , Mo 2+ , Mn 2+ , Mg 2+ , Ni 2+ <strong>and</strong> Zn 2+ . Many ions,<br />
particularly those of heavy metals, are toxic in excess. Metal ions such as Na + ,in high<br />
concentrations, exert asalt stress on yeast (Section 12.3.1). Precise requirements are<br />
strain specific <strong>and</strong> the combination <strong>and</strong> concentration of mineral ions available in the<br />
medium is important since synergistic <strong>and</strong> antagonistic interactions occur. Some metal<br />
ions may be tolerated by certain yeast strains but be growth inhibitory to others. For<br />
example, a commonly used test to differentiate between brewing <strong>and</strong> non-brewing `wild'<br />
yeast strains is based on the ability of the latter to grow in the presence of relatively high<br />
concentrations of copper ions. A few ions, notably Mg 2+ <strong>and</strong> K + , are required at higher<br />
concentrations, typically at the millimolar level.<br />
Brewers' malt wort supplies all the mineral nutritional requirements of yeast, with the<br />
possible exception of zinc. Zinc ions can be chelated by wort amino acids, proteins <strong>and</strong><br />
phytate <strong>and</strong> a proportion of these may be removed as insoluble precipitates during the<br />
copper boil (Daveloose, 1987). For this reason, zinc supplements are commonly added to<br />
wort in the fermenter.<br />
12.3.5 Growth factors<br />
Growth factors are a diverse group of organic compounds which individual yeast strains<br />
are unable to synthesize <strong>and</strong> which are essential for growth. Their presence in the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
medium is obligatory. The group includes vitamins, some purines, pyrimidines,<br />
polyamines, nucleosides, nucleotides <strong>and</strong> certain lipids. Growth factors may be required<br />
as intermediates in certain essential metabolic pathways or have astructural role. More<br />
commonly, they are essential catalytic components of coenzymes as exemplified by<br />
vitamins. Concentrations required for growth are low, typically in the M range.<br />
Vitamins, or derivatives of them, are involved in many fundamental biochemical<br />
processes, for example, biotin (carboxylation reactions), thiamine (oxo-acid decarboxylation<br />
reactions), nicotinic acid (redox reactions), pyridoxine (transaminations), paminobenzoic<br />
acid (one-carbon transfer) <strong>and</strong> pantothenic acid (acetylation reactions)<br />
(Fig. 4.29). The requirement for vitamins <strong>and</strong> other growth factors varies widely between<br />
individual strains. Individual strain requirements have taxonomic significance (Kregervan<br />
Rij, 1984; Kurtzman <strong>and</strong> Fell, 1988).<br />
With regard to brewing strains of S. cerevisiae, all require biotin <strong>and</strong> pantothenic<br />
acid. Many strains require inositol <strong>and</strong> thiamine. In general, lager strains have agreater<br />
requirement for growth factors compared to ale strains. Most brewers' worts contain<br />
adequateconcentrationsofgrowthfactorsforallyeaststrains.Inrarecases,anutritional<br />
supplement, enriched in growth factors, is added to worts that are considered deficient.<br />
Such supplements, termed `yeast foods', are partially purified extracts of yeast,<br />
occasionally with added vitamins. Under anaerobic conditions Saccharomyces strains<br />
become auxotrophic (have aabsolute requirement for growth) for certain lipids. Thus,<br />
some sterols <strong>and</strong> unsaturated fatty acids, accumulated during prior aerobic growth or<br />
supplied in the medium, are essential components of membranes. In addition,<br />
unsaturated fatty acids appear to have roles as regulatory effector molecules. The<br />
syntheses of sterols <strong>and</strong> unsaturated fatty acids, de novo, require molecular oxygen<br />
(Section 12.7).<br />
12.4 Nutrient uptake<br />
Yeasts possess mechanisms to regulate the passage of nutrients from the external medium<br />
into the cell (Cartwright et al., 1989). The plasma membrane forms the principal semipermeable<br />
barrier through which all nutrients must pass. Cells have systems for sensing<br />
the nature <strong>and</strong> concentration of nutrients in the external medium. They have the ability to<br />
selectively assimilate individual compounds from complex mixtures in an ordered<br />
manner. Some nutrients can be transported into the cell against a concentration gradient.<br />
In addition to assimilating compounds from the external medium, some of the products of<br />
metabolism are excreted from the cell.<br />
Cartwright <strong>and</strong> co-workers (1989) differentiate between vectoral <strong>and</strong> scalar<br />
metabolism. The latter is that portion of metabolism in which molecules are subject to<br />
chemical modification with no appreciable movement. The former is defined as that<br />
portion of metabolism that involves the controlled physical movement of molecules.<br />
Some vectoral metabolism is intracellular, for example, the controlled transport of<br />
metabolites between intracellular compartments. Although it is convenient to consider<br />
these two facets of metabolism as separate entities, they are, of course, regulated <strong>and</strong> coordinated<br />
processes.<br />
A number of distinct mechanisms result in the passage of solutes across membranes.<br />
The simplest <strong>and</strong> slowest is free diffusion in which molecules traverse the membrane via<br />
the lipid components driven by a concentration gradient. Transport ceases when the<br />
solute concentration on each side of the membrane becomes equal. Other solute transport<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
processes require the intervention of specific membrane proteins, termed transporters or<br />
permeases. Facilitated diffusive transporters regulate the passage of solutes in the<br />
direction of a concentration gradient. As in the case of free diffusion, transport ceases<br />
when the concentration gradient ceases to exist.<br />
The most prevalent type of transport system in yeast is that in which solutes passage<br />
is against a concentration gradient. The process is termed active transport since it<br />
requires metabolic energy. Movement of individual solutes is controlled by specific<br />
permeases, which may be inducible or constitutive. The energy is provided by the<br />
plasma membrane H + -ATPase (Serrano, 1989). The latter energizes the membrane by<br />
promoting the uni-directional movement of protons utilizing the energy released by the<br />
hydrolysis of ATP. The resultant proton motive force is a combination of the membrane<br />
potential <strong>and</strong> the proton gradient. Solute molecules usually travel across the membrane<br />
accompanied by protons, a mechanism termed proton symport transport. Occasionally,<br />
solute movement occurs against a reverse passage of protons, termed proton antiport<br />
transport.<br />
Commonly, yeast cells possess multiple carriers for the same nutrient, or class of<br />
nutrients. Frequently, these carriers have different affinities for the same substrate. Thus,<br />
within the same strain there may be both high <strong>and</strong> low affinity transporters for given<br />
substrates. Often the latter is constitutive <strong>and</strong> the former inducible. Presumably, the high<br />
affinity inducible systems represent an evolutionary mechanism, which confers a<br />
selective advantage where mixed populations of yeast <strong>and</strong> other micro-organisms are<br />
competing for small amounts of essential nutrients.<br />
Yeast membranes, including the plasma membrane <strong>and</strong> others enclosing intracellular<br />
organelles, possess channels which have been implicated in the transport of ions, water<br />
<strong>and</strong> some organic molecules, such as glycerol (Gustin et al., 1986; Luyten et al., 1995).<br />
The channels are proteinaceous in nature <strong>and</strong> are activated or deactivated by<br />
perturbations in membrane polarization. The predominant ion channel controls the<br />
efflux of K + . This phenomenon balances the influx of protons that accompany sugar<br />
uptake <strong>and</strong> is, therefore, necessary for intracellular charge homeostasis. Solute molecules<br />
may be transported enclosed in a vesicle, which arises from the plasma membrane. This<br />
process is termed pinocytosis. Its occurrence in yeast is disputed, although it may be<br />
implicated in the intracellular trafficking of macromolecules.<br />
The transport systems described so far facilitate the entry of solutes without chemical<br />
modification. Group translocation systems mediate the uptake of solutes <strong>and</strong> in so doing<br />
modify chemical structures. In yeast, glucose uptake is coupled to phosphorylation. This<br />
process may be associated with the glucose transporters (Lagunas, 1993).<br />
Before solutes pass through the plasma membrane <strong>and</strong> into the cell, they must<br />
negotiate the capsule, if present, the cell wall <strong>and</strong> the intervening periplasm. Yeast cell<br />
walls are freely permeable but their physical structure places an upper limit on the size of<br />
molecules that can pass through unimpeded. This feature is of benefit in that it allows<br />
retention of enzymes in the periplasm. The upper limit of cell wall porosity is in the<br />
region of 200±400 kDa (de Nobel et al., 1991). The comparatively wide range reflects<br />
changes in cell wall porosity that occur during different phases in the yeast cell growth<br />
cycle.<br />
12.4.1 Sugar uptake<br />
The uptake of sugars by Saccharomyces strains have been subject to the closest scrutiny<br />
as befits their role in industrial fermentations (Andre, 1995; Horak, 1997). Although<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
g l –1<br />
g l –1<br />
10<br />
8<br />
6<br />
4<br />
2<br />
100<br />
80<br />
60<br />
40<br />
20<br />
Glucose<br />
0<br />
0 20 40 60 80 100<br />
Time (h)<br />
Maltose<br />
Maltotriose<br />
Fructose<br />
Sucrose<br />
0<br />
0 20 40 60 80 100<br />
Time (h)<br />
Fig. 12.4 Patterns of uptake of fermentable sugars from 10 ëPlato ale wort. The yeast was a top<br />
fermenting ale strain <strong>and</strong> the temperature was 18 ëC (C. A. Boulton, unpublished data).<br />
some sugars may enter the cell by free diffusion, uptake appears to be predominantly via<br />
active processes <strong>and</strong> is against aconcentration gradient. Control of sugar uptake is<br />
complex <strong>and</strong> highly regulated. In industrial fermentations, using feedstocks containing<br />
complex mixtures of several sugars, uptake into the cell limits the overall rate of ethanol<br />
formation.<br />
When presented with amixture of assimilable sugars, yeasts have mechanisms for<br />
selecting first, those which are most readily utilized. In the case of brewers' wort, the<br />
utilization of sugars is an ordered process (Fig. 12.4). Sucrose is hydrolysed by an<br />
invertase that is secreted into the periplasm. This results in atransient increase in the<br />
concentrations of fructose <strong>and</strong> glucose. Fructose <strong>and</strong> glucose are assimilated<br />
simultaneously. The predominant sugar, maltose is then taken up. When the maltose<br />
concentration falls to an undetectable level maltotriose is assimilated. Longer chain<br />
sugars are not utilized by brewing yeasts.<br />
Amultiplicity of hexose uptake systems has been identified in various yeast strains<br />
(Kruckenberg, 1996; Ozcan <strong>and</strong> Johnston, 1999). Glucose is the preferred substrate but<br />
frequentlyotherhexoses are also transported.In S.cerevisiae, atleast 19separate genes are<br />
responsible for the synthesis of hexose carriers. Two classes of hexose carrier are<br />
recognized,termedhigh<strong>and</strong>lowaffinity.ThehighaffinitysystemstransportD-glucose,Dfructose<strong>and</strong>D-mannose<strong>and</strong>areofthefacilitateddiffusionvariety.TheyhaveK<br />
mvaluesof<br />
approximately 1mM. In the presence of relatively high glucose concentrations (>0.1M)<br />
activityisrepressed(Does<strong>and</strong>Bisson,1989).Thisphenomenonispartofawidersystemof<br />
metabolic control, termed catabolite repression (Section 12.5.8). Activity of the high<br />
affinity carriers is associated with phosphorylation of the substrate via the glycolytic<br />
hexokinases <strong>and</strong> a glucokinase. Whether, or not, the phosphorylation involves interactions<br />
between kinases <strong>and</strong> the glucose carrier remains to be elucidated (Lagunas, 1993).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The low affinity hexose carriers are constitutive <strong>and</strong> have Km values in the region of<br />
20mM. Their existence has been disputed since it has been claimed that low affinity<br />
glucose uptake could simply reflect passive diffusion. However, this suggestion has<br />
been refuted on the basis that uptake rates are 2±3 times higher than would be predicted<br />
for passive diffusion (Gamo et al., 1995). The reason(s) for the multiplicity of hexose<br />
carriers is not clear. Detailed genetic analysis is needed to determine under what<br />
conditions each is active. It has been demonstrated that some of the carriers are subject<br />
to nitrogen catabolite inactivation (Busteria <strong>and</strong> Lagunas, 1986; Section 12.8). This<br />
process occurs following the exhaustion from the medium of certain nitrogen sources<br />
<strong>and</strong> leads to inactivation of hexose <strong>and</strong> other sugar uptake systems via proteolysis of<br />
the carriers. Probably other global metabolic control mechanisms, such as the<br />
availability of oxygen <strong>and</strong> other nutrients, will also influence the activity of individual<br />
hexose carriers.<br />
Although the glucose transporters show activity towards other hexoses, S. cerevisiae<br />
also possesses a specific transport system for galactose. In common with the glucose<br />
uptake system, both constitutive low affinity <strong>and</strong> inducible high affinity galactose<br />
transporters exist. There are specific uptake systems for pentoses, in those yeast strains<br />
capable of utilizing them. In general, the activity of these is repressed by glucose.<br />
Sucrose assimilation is dependent on the production of invertase. Synthesis of this<br />
enzyme is dependent on the presence of a number of SUC genes (SUC1±SUC5 <strong>and</strong><br />
SUC7). Yeast cells produce an intracellular constitutive invertase, whose function is<br />
unknown. A second invertase is secreted into the periplasm <strong>and</strong> this is responsible for the<br />
assimilation of extracellular sucrose. The enzyme is also active towards raffinose. The<br />
fructose <strong>and</strong> glucose produced by the hydrolysis of sucrose are both transported via the<br />
hexose carrier system. The periplasmic invertase is subject to glucose repression<br />
although, surprisingly, for maximum transcription of SUC2 to occur a low concentration<br />
(0.1%) of glucose is required (Ozcan et al., 1997).<br />
Specific disaccharide carriers mediate the uptake of maltose (4.4) <strong>and</strong> trehalose ( -Dglucopyranosyl<br />
(1, 1)- -D-glucopyranose). A constitutive low affinity transport system<br />
<strong>and</strong> a high affinity proton symport carrier accommodate trehalose uptake. The latter is<br />
subject to glucose repression. Derepressed yeast cells have high levels of intracellular<br />
trehalose. This suggests that under these conditions trehalose accumulation is against a<br />
concentration gradient. Probably the cell does this since there is evidence that trehalose 6phosphate<br />
is involved in the control of glycolysis (Thevelein <strong>and</strong> Hohmann, 1995).<br />
Maltose is the most abundant sugar in wort. Its uptake <strong>and</strong> utilization is controlled by a<br />
complex series of MAL genes. These occur at five unlinked homologous loci, not<br />
restricted to a single chromosome. Each locus consists of three genes, which encode for a<br />
maltose carrier, maltase <strong>and</strong> a post-transcriptional regulator of the carrier <strong>and</strong> maltase<br />
genes. Maltose utilization is repressed by glucose <strong>and</strong> requires the maltose for induction<br />
(Busteria <strong>and</strong> Lagunas, 1986). Maltose utilization is subject to nitrogen catabolite<br />
inactivation. Thus, under conditions of nitrogen exhaustion or in the presence of glucose,<br />
the maltose permease is irreversibly inactivated via the action of a protease. Maltose<br />
uptake is an energy-requiring proton symport process. Efflux of K + ensures<br />
electrochemical neutrality. The uptake system is of the high affinity variety. In brewing<br />
strains of S. cerevisiae, a second low affinity system has been identified (Crumplen et al.,<br />
1996). Transport of maltotriose in S. cerevisiae is accomplished by a constitutive<br />
facilitated diffusion carrier. It has been suggested that the permease may be absent in<br />
some ale strains thus accounting for the observations that some of the latter are not able to<br />
utilize maltotriose (Stewart et al., 1995).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
12.4.2 Uptake of nitrogenous nutrients<br />
Yeasts possess transport systems for mediating the uptake of both inorganic <strong>and</strong> organic<br />
nitrogen sources. In wort, yeast is presented with acomplex mixture of nitrogen sources.<br />
As is the case with the utilization of carbon sources, uptake of nitrogenous nutrients is an<br />
ordered process. Thus, the presence in the medium of readily assimilable nitrogen sources<br />
represses the synthesis of the uptake systems <strong>and</strong> catabolic enzymes of other less readily<br />
utilized sources of this nutrient. This is termed nitrogen catabolite repression (Wiame et<br />
al., 1985).<br />
All yeasts can utilize ammonium ions <strong>and</strong> indeed this nutrient is usually utilized in<br />
preference to organic sources of nitrogen. However, in brewing yeast, during<br />
fermentation, several amino acids are utilized before ammonium ions. In S. cerevisiae<br />
high <strong>and</strong> low affinity carriers control the uptake of ammonium ions. The high affinity<br />
carrier is an active transport system <strong>and</strong> requires the presence of an oxidizable substrate<br />
for activity. S. cerevisiae does not utilize nitrate or nitrite. Individual yeast strains can<br />
assimilate awide range of organic sources of nitrogen. Most strains can utilize urea <strong>and</strong><br />
in S. cerevisiae high <strong>and</strong> low affinity urea transporters occur. Anumber of transporters<br />
occur in yeast specific for one or small groups of amino acids. In addition, there is a<br />
general amino acid permease (GAP) with broad specificity. Some of the transporters,<br />
including GAP, are repressed by ammonium ions, asparagine <strong>and</strong> glutamine (Grenson,<br />
1992). In S. cerevisiae, 12 constitutive <strong>and</strong> four nitrogen-repressible amino acid carriers<br />
have been identified (Horak, 1986). The transporters can be of the high or low affinity<br />
type, uptake is active <strong>and</strong> involves proton symport.<br />
Maximum activity of GAP occurs only under conditions of nitrogen starvation. In this<br />
regard, GAP functions as anitrogen scavenger. Regulation of theother carriers is complex<br />
<strong>and</strong> dependent on the spectrum <strong>and</strong> concentrations of amino acids present in the medium.<br />
The presence of multiple carriers for amino acids affords the yeast an opportunity to order<br />
uptake in response to need. Based upon chemostat studies with S. cerevisiae growing<br />
underconditionsofnitrogenlimitation, ithasbeensuggested(Oliveraetal.,1993)thatthe<br />
specific permeases are involved in the uptake of amino acids destined for use in anabolic<br />
pathways.Conversely,thosepermeasessubjecttonitrogencataboliterepression,including<br />
GAP, mediate the uptake of amino acids used in catabolic pathways.<br />
In brewing yeasts growing on wort, the uptake of amino acids is an ordered process.<br />
Pierce (1987) divided amino acids into four classes, based on their order of assimilation<br />
from wort during fermentation (Table 12.2). Surprisingly, in view of its role in nitrogen<br />
catabolite repression, ammonia is not a member of the first group, although asparagine<br />
<strong>and</strong> glutamine are. The amino acids in classes A <strong>and</strong> B are required for anabolic<br />
metabolism, principally protein synthesis. They are taken up by those permeases that are<br />
not subject to nitrogen catabolite repression. Conversely, those in Class C are only taken<br />
up when the Class A amino acids have disappeared <strong>and</strong> nitrogen catabolite repression is<br />
relieved. Proline is the sole member of class D. This imino acid is not utilized during<br />
brewery fermentation since its oxidation requires a mitochondrial oxidase, which is<br />
repressed during fermentation (Wang <strong>and</strong> Br<strong>and</strong>riss, 1987).<br />
Short homopeptides may be taken up by yeast although not as readily as the free<br />
amino acids (Ingledew <strong>and</strong> Patterson, 1999). Ammonia inhibits the uptake of dipeptides.<br />
Peptides containing no more than five amino acid residues may be transported into the<br />
cell. The carrier in S. cerevisiae is reportedly of broad specificity, active <strong>and</strong> capable of<br />
transporting di- <strong>and</strong> tripeptides (Marder et al., 1977). S. cerevisiae strains are not capable<br />
of transporting oligopeptides <strong>and</strong>, since they do not produce exogenous proteases, are not<br />
able to utilize these nitrogen sources.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 12.2 Amino acid classification based on the order of assimilation from wort during<br />
fermentation (Pierce, 1987). Amino acids are assimilated in the order A, B, C, D<br />
Class A Class B Class C Class D<br />
Arginine Histidine Alanine Proline<br />
Asparagine Isoleucine Ammonia<br />
Aspartate Leucine Glycine<br />
Glutamate Methionine Phenylalanine<br />
Glutamine Valine Tryptophan<br />
Lysine Tyrosine<br />
Serine<br />
Threonine<br />
Uptake of purines <strong>and</strong> pyrimidines is accommodated by both active proton symport<br />
systems <strong>and</strong> by facilitated diffusion carriers. The active systems reportedly mediate the<br />
uptake of adenine, adenosine, cytosine, guanine <strong>and</strong> hypoxanthine. Transport of uracil<br />
<strong>and</strong> uridine is apparently by facilitated diffusion (Cartwright et al., 1989).<br />
12.4.3 Lipid uptake<br />
Saccharomyces cerevisiae utilizes lipids such as fatty acids <strong>and</strong> sterols. These may be<br />
used for direct incorporation into cellular structures, as sources of metabolic<br />
intermediates for both catabolic <strong>and</strong> anabolic pathways or to fulfil roles in cellular<br />
signalling systems. At high concentrations, fatty acids are taken up by simple diffusion, a<br />
process aided by the lipophilic nature of the plasma membrane (van der Rest et al., 1995).<br />
In S. cerevisiae, the presence of a medium chain length fatty acid transporter has been<br />
inferred (Faergeman et al., 1997). In anaerobic brewery fermentations, brewing yeasts are<br />
auxotrophic for unsaturated fatty acids. These compounds must be synthesized during the<br />
aerobic phase of fermentation. Some of this requirement may be satisfied by direct uptake<br />
from wort.<br />
Yeast can assimilate exogenous sterols but only under anaerobic conditions when de<br />
novo synthesis is precluded. This phenomenon is termed aerobic sterol exclusion. Sterol<br />
uptake requires expression of the SUT1 <strong>and</strong> SUT2 genes (Bourdot <strong>and</strong> Karst, 1995).<br />
These are hypoxic genes, expressed only under anaerobic conditions. It is assumed that<br />
the product of their expression inhibits the transcription of the genes encoding the sterol<br />
transporter.<br />
12.4.4 Ion uptake<br />
Uptake of metal ions by yeast is a biphasic process. Firstly, ions are concentrated by<br />
attachment to the cell surface, a passive process termed biosorption. Suggested<br />
mechanisms for attachment to the cell wall include complexation, ion exchange,<br />
adsorption <strong>and</strong> precipitation (Blackwell et al., 1995). The process is independent of<br />
temperature, does not require metabolic energy or indeed viability. It has been suggested<br />
that it may serve as a protective strategy for the removal of potentially toxic ions.<br />
Secondly, ions are transported across the plasma membrane <strong>and</strong> into the cell by<br />
bioaccumulation. This is an active process involving proton symport <strong>and</strong> K + efflux. Once<br />
in the cell, metal ions are commonly compartmentalized in the vacuolar system. Metal<br />
ion uptake is tightly regulated since individual ions may be essential or toxic at low <strong>and</strong><br />
high concentration, respectively. Different yeast species possess metal ion carriers of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
oth broad <strong>and</strong> narrow specificity. Metal ion transport is a dynamic process <strong>and</strong><br />
synergistic <strong>and</strong> antagonistic effects are possible depending on the spectrum of ions<br />
present. Metal ion uptake <strong>and</strong> efflux occur reflecting the need of the cell to maintain an<br />
internal ionic balance <strong>and</strong> still supply ions where needed for proper enzyme function<br />
(Jones <strong>and</strong> Gadd, 1990).<br />
S. cerevisiae possesses a metal ion carrier with broad specificity, capable of<br />
transporting the divalent ions of calcium, cobalt, nickel, manganese, magnesium,<br />
strontium <strong>and</strong> zinc. The carrier has the highest affinity for Mg 2+ <strong>and</strong> uptake of this ion<br />
occurs simultaneously with phosphate. Uptake of Co 2+ ,Mn 2+ ,Fe 2+ <strong>and</strong> Zn 2+ ions in S.<br />
cerevisiae is mediated by distinct high <strong>and</strong> low affinity transporters. Activity of the<br />
individual carriers is dependent upon the availability of the ions <strong>and</strong> typically, the high<br />
affinity permeases have scavenging roles. Uptake of manganese is competitively<br />
inhibited by magnesium ions. The iron transporter also mediates the uptake of cadmium,<br />
cobalt <strong>and</strong> nickel.<br />
Zincuptakeisofparticularrelevancetobrewinginthatmalt worts maybedeficient in<br />
this ion (Daveloose, 1987; Chapter 4). It plays an essential role in the function of many<br />
enzymes, including alcohol dehydrogenase. S. cerevisiae maintains zinc homeostasis by<br />
the operation of carriers that mediate both uptake <strong>and</strong> efflux. Copper is an essential<br />
nutrient at low concentration but can be toxic at higher levels. The ability of certain<br />
strains of `wild' yeast to grow in the presence of copper at a concentration that brewing<br />
strains of S. cerevisiae cannot tolerate is of diagnostic importance. Copper uptake is<br />
mediated by a high affinity carrier expressed by the CTR 1 gene. The transporter may<br />
also be responsible for the uptake of ferrous ions. Copper homeostasis is achieved by<br />
regulated intracellular sequestration in the form of a metalothionin protein. In addition,<br />
transporters are present that function as mediators of copper ion efflux.<br />
Calcium ions have important roles in metabolic signalling systems. They are<br />
implicated in the control of the mating response (not applicable to brewing yeast),<br />
modulation of cellular growth <strong>and</strong> progress through the cell cycle. In consequence,<br />
intracellular calcium ion concentration is highly regulated (Youatt, 1993). An active<br />
proton antiport permease regulates calcium transport across both the plasma <strong>and</strong> vacuolar<br />
membranes. Calcium ions are sequestered in the cell by specific binding proteins such as<br />
calmodulin.<br />
Transport of potassium ions is used by yeast cells to maintain charge homeostasis.<br />
Frequently, efflux of potassium ions is used by yeast in conjunction with proton symport<br />
permeases, to maintain electrochemical neutrality. It is the most common intracellular ion<br />
in the yeast cytosol. The high intracellular concentration is maintained by several parallel<br />
transport systems. These include uptake via both active systems <strong>and</strong> plasma membrane<br />
pores. In S. cerevisiae, the active system is a proton antiporter.<br />
Phosphate uptake is dependent on the presence of a fermentable substrate <strong>and</strong> both<br />
high <strong>and</strong> low affinity transporters occur in S. cerevisiae. Under conditions of phosphate<br />
limitation, the PHO5 gene is induced. This encodes for a secreted acid phosphatase <strong>and</strong><br />
this is involved in phosphate scavenging. In some yeasts, for example Zygosaccharomyces<br />
bailii, active carriers exist for the uptake of sulphate <strong>and</strong> sulphite. In brewing<br />
yeast, uptake of the latter is apparently via simple diffusion of sulphur dioxide, whereas<br />
several carriers mediate the uptake of sulphate. The presence of sulphite inhibits the<br />
utilization of sulphate.<br />
Movement of hydrogen ions into <strong>and</strong> out of the cell is of paramount importance in<br />
controlling the transport of other charged species <strong>and</strong> maintenance of intracellular pH.<br />
The trans-membrane proton gradient <strong>and</strong> proton-motive force is generated by<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
H + -ATPase. The importance of the latter is indicated by the fact that it is the most<br />
abundant membrane protein (Serrano, 1989).<br />
12.4.5 Transport of the products of fermentation<br />
During fermentation, the transformation of wort into beer is accompanied by adecrease<br />
in pH. Much of this decrease is aresult of the proton antiport component of other uptake<br />
systems. Acidification of beer is also contributed to by the formation of carbonic acid<br />
derived from the carbon dioxide produced during fermentation. In addition, several<br />
organic acids, notably, lactic, citric, pyruvic, malic, acetic, formic, succinic <strong>and</strong> butyric<br />
acids are excreted from fermenting yeast cells. In the latter stages of fermentation, some<br />
of these compounds may be re-assimilated. Transport may be via simple diffusion or<br />
active uptake systems involving proton symport.<br />
Many products of yeast metabolism contribute to beer flavour (Chapter 23). These<br />
include ethanol, higher alcohols, esters, aldehydes <strong>and</strong> vicinal diketones. It follows that<br />
these metabolic by-products must be transported out of or released from the cell during<br />
fermentation. This subject has received little attention; probably non-concentrative<br />
diffusion is the mechanism used.<br />
12.5 Sugar metabolism<br />
The catabolism of sugars provides yeast with energy <strong>and</strong> carbon skeletons for anabolic<br />
pathways. This is an essential activity <strong>and</strong> in consequence, alarge proportion of total<br />
metabolism is devoted to it. Several distinct pathways are involved. The flux of carbon<br />
through individual pathways is influenced by yeast genotype <strong>and</strong> its phenotypic<br />
expression as brought about by the conditions to which they are exposed.<br />
12.5.1 Glycolysis<br />
Glycolysis, or the Embden-Myerhof-Parnas pathway, is the major sugar catabolic<br />
pathway in yeast. It operates under both aerobic <strong>and</strong> anaerobic conditions <strong>and</strong> is the route<br />
by which approximately 70% of exogenous hexose sugars are assimilated. The<br />
importance of this pathway is reflected in the fact that glycolytic enzymes comprise<br />
30±65% of the total soluble protein pool in S. cerevisiae. (Fraenkel, 1982). The pathway<br />
catalyses the conversion of one molecule of glucose into two molecules of pyruvate (Fig.<br />
12.5). The initial phosphorylation reaction, in which ATP is the phosphate donor, may be<br />
catalysed by one of three enzymes. Hexokinases 1<strong>and</strong> 2show activity towards both<br />
glucose <strong>and</strong> fructose <strong>and</strong> glucokinase with glucose, alone. All show activity towards<br />
mannose. All of the glycolytic reactions are reversible with the exceptions of the initial<br />
phosphorylation ofglucose,the phosphorylation offructose-6-phosphate toyieldfructose<br />
1,6bisphosphate <strong>and</strong> the dephosphorylation of phospho-enol-pyruvate to form pyruvate.<br />
Several of the steps are catalysed by multiple enzymes, as indicated. Glycolysis can<br />
operate in the reverse (gluconeogenic) direction. In this case, three additional enzymes,<br />
phospho-enol-pyruvate carboxykinase, fructose 1,6-bisphosphatase <strong>and</strong> glucose 6phosphate<br />
phosphatase catalyse the contra-flow of carbon past the irreversible steps of<br />
glycolysis. Other sugars feed into the glycolytic pathway as shown in Fig. 12.6. Like<br />
glucose, the utilization of these other sugars also involves reactions in which ATP is<br />
consumed <strong>and</strong> phosphorylated intermediates are formed. Some of the reactions use the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HXK1<br />
HXK2<br />
GLK1<br />
PGI1<br />
PFK1<br />
PFK2<br />
CH2OH O<br />
H H<br />
H<br />
HO OH H OH<br />
H OH<br />
Dihydroxyacetone<br />
phosphate<br />
Oxaloacetate<br />
Glucose<br />
Glucose 6-phosphate<br />
CH2O P<br />
H<br />
O<br />
H<br />
H<br />
HO HO<br />
OH<br />
H OH<br />
Fructose 6-phosphate<br />
P OCH2 CH2OH O<br />
H HO<br />
H OH<br />
OH H<br />
CH2O P CO CH2OH Fructose 1,6-bisphosphate<br />
FBP1<br />
PCK1<br />
TPI1<br />
CH2O P CHOH COO P<br />
CH2O CHOH COO –<br />
P<br />
CH2OH COO –<br />
CHO P<br />
Glyceraldehyde<br />
3-phosphate<br />
1,3-Diphosphoglycerate<br />
3-Phosphoglycerate<br />
2-Phosphoglycerate<br />
Phosphoenolpyruvate<br />
Pyruvate<br />
coenzyme, uridine triphosphate (UTP), which gives rise to acarrier of sugar molecules<br />
(e.g. UDPG; 4.55).<br />
Glycolysis generates reducing power in the form of NADH. This is re-oxidized in<br />
redox balancing reactions (Section 12.2). During the conversion of glucose to fructose<br />
1, 6-bisphosphate, two molecules of ATP are consumed. In the later stages of glycolysis,<br />
four ATP molecules are generated in the reactions catalysed by phosphoglycerokinase<br />
ATP<br />
ADP<br />
ATP<br />
ADP<br />
P OCH2 CH2O P<br />
O<br />
H HO<br />
H OH<br />
OH H<br />
GPP<br />
FBP1<br />
CH2O P CHOH CHO<br />
COO –<br />
CH2 CO P<br />
CH 3 CO<br />
COO –<br />
NAD +<br />
NADH<br />
ADP<br />
ATP<br />
ADP<br />
ATP<br />
TDH1<br />
TDH2<br />
TDH3<br />
PGK1<br />
GPM1<br />
ENO1<br />
ENO2<br />
PYK1<br />
PYK2<br />
Fig. 12.5 The main glycolytic pathway. The genes <strong>and</strong> enzymes resposible for each step are:<br />
HXK1, HXK2, hexokinases; GLK1, glucokinase; PGI1, phosphoglucose isomerase; PFK1, PFK2,<br />
phosphofructokinase; FBP1, fructose 1, 6-bisphosphatase; FBA1, fructose 1, 6-bisphosphate<br />
aldolase; TPI1, triose phosphate isomerase; TDH1, TDH2, TDH3, glyceraldehyde 3-phosphate<br />
dehydrogenase; PGK1, phosphoglycerate kinase; GPM1, glycerophosphate mutase; ENO1, ENO2',<br />
enolase; PYK1, PYK2, pyruvate kinase. Three gluconeogenic enzymes are also shown: GPP,<br />
glucose 6-phosphate phosphatase; FBP1; fructose 1, 6-bisphosphatase; PCK1, phosphoenolpyruvate<br />
carboxykinase.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Sucrose<br />
Maltose<br />
(4.4)<br />
Glucose<br />
Glucose 6-phosphate<br />
Fructose 6-phosphate<br />
GLYCOLYSIS<br />
Fructose (4.3)<br />
Trehalose<br />
Trehalose<br />
phosphate<br />
Mannose<br />
phosphate<br />
Galactose (4.9)<br />
Galactose<br />
1-phosphate<br />
UDP-glucose<br />
Glucose 1-phosphate<br />
Melibiose<br />
Raffinose<br />
Mannose<br />
(4.8)<br />
UDP-galactose<br />
Fig. 12.6 The mode of entry of various sugars into the glycolytic pathway. The numbers in<br />
parentheses indicate where the molecular structures may be found.<br />
<strong>and</strong> pyruvate kinase. Therefore, for every molecule of glucose catabolized there is anet<br />
gain of two molecules of ATP. This phenomenon is termed substrate level<br />
phosphorylation. It is the predominant mechanism used by yeast to generate energyrich<br />
compounds under fermentative conditions. Glycolysis is active in yeast under all<br />
conditions <strong>and</strong> the component enzymes are constitutive. During growth on sugars, the<br />
direction of carbon flow is from glucose to pyruvate. During growth on oxidative carbon<br />
sources such as ethanol, the glycolytic pathway is reversed <strong>and</strong> is used to generate<br />
intermediates for anabolism. In this respect, glycolysis/gluconeogenesis is an amphibolic<br />
pathway, which serves both anabolic <strong>and</strong> catabolic roles. Reverse glycolysis is part of<br />
gluconeogenic sugar generating metabolism (Section 12.5.6).<br />
Carbon flux through glycolysis is aregulated process <strong>and</strong> controls are exerted on both<br />
gene expression <strong>and</strong> enzyme activity. Transport of sugars into the cell, phosphorylation<br />
of glucose <strong>and</strong> regulation of the activities of phosphofructokinase <strong>and</strong> pyruvate kinase by<br />
metabolic effectors have all been implicated. Possibly uptake of sugars is the ratedetermining<br />
step in glycolysis since the maximum rates of transport are close to the<br />
maximum observed rates of glycolytic flux. The intracellular concentration of glucose is<br />
always lower than that in the external medium. Nitrogen starvation, which brings about a<br />
progressive decline in rates of glycolytic flux, affects rates of sugar transport but not the<br />
activities of the glycolytic enzymes.<br />
Modulation of the activity of hexokinases by trehalose 6-phosphate may be of<br />
regulatory significance. Trehalose 6-phosphate is a strong competitive inhibitor of<br />
hexokinase <strong>and</strong> this may be used to control the entry of glucose into glycolysis.<br />
Alternatively, or possibly as well as, the synthesis of trehalose (Section 12.5.7) may be<br />
used as a mechanism for controlling levels of phosphate. During exponential growth of S.<br />
cerevisiae on glucose, the maximum glycolytic flux is within the range 200±<br />
300 mol.hexose/min./g dry weight of yeast. Most of the glycolytic enzymes, under<br />
optimal conditions, are capable of greater activity than this, indicating that they are<br />
probably not rate-determining. Phosphofructokinase alone has a maximum activity close<br />
to the measured maximum rate of glycolytic flux, suggesting a possible regulatory role.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Further evidence for this regulatory role is the fact that the activity of this enzyme is<br />
modulated by several effectors. Phosphofructokinase has an obligate requirement for<br />
Mg 2+ .NH4 + <strong>and</strong> K + .Binding of fructose 6-phosphate is positively co-operative. ATP<br />
inhibits activity <strong>and</strong> AMP is stimulatory.<br />
The concerted effect of these metabolites on the activity of phosphofructokinase<br />
possibly explains why in yeast, under some conditions, flux through glycolysis becomes<br />
oscillatory. The explanation for this also supports the rate-determining role of<br />
phosphofructokinase in glycolysis. The oscillatory behaviour can be induced in cellfree<br />
extracts by the addition of fructose 6-phosphate but not fructose 1,6-bisphosphate,<br />
indicating that phosphofructokinase is the site of the effect. Fructose 6-phosphate<br />
activatesphosphofructokinase,which results inadecline inthe concentration of ATP<strong>and</strong><br />
concomitant increase in levels of ADP <strong>and</strong> AMP. In turn, this activates phosphoglycerate<br />
kinase <strong>and</strong> pyruvate kinase further down the glycolytic pathway. In addition, fructose<br />
1,6-bisphosphate activates pyruvate kinase. This removes ADP <strong>and</strong> AMP <strong>and</strong> increases<br />
the concentration of ATP via substrate level phosphorylation. High levels of ATP <strong>and</strong><br />
reduced fructose 6-phosphate now combine to reduce the activity of phosphofructokinase.<br />
In consequence, high ATP concentration inhibits the glycolytic kinases, fructose 6phosphate<br />
accumulates <strong>and</strong> the cycle is triggered again.<br />
Modulation of the activity of phosphofructokinase by substrates <strong>and</strong> products of the<br />
reaction undoubtedly has significance. However, in terms of overall control of glycolytic<br />
rates, the effect of another metabolite, fructose 2,6-bisphosphate appears to be of greater<br />
importance. Fructose 2,6-bisphosphate (F2,6bP) is produced from fructose 6-phosphate<br />
<strong>and</strong> ATP by the action of 6-phosphofructo-2-kinase (6PF2K). In yeast, asecond enzyme,<br />
fructose 2,6-bisphosphatase (2,6bPase) degrades F2,6bP to fructose 6-phosphate <strong>and</strong><br />
phosphate. Fructose 6-phosphate is avery potent activator of 6-phosphofructo-1-kinase,<br />
the major glycolytic enzyme. Furthermore, F2,6bP inhibits the activity of the<br />
gluconeogenic enzyme, fructose 1,6-bisphosphatase.<br />
The activities of 6PF2K <strong>and</strong> 2,6bPase are regulated by reversible phosphorylation. In<br />
the phosphorylated form, 6PF2K is activated <strong>and</strong> 2,6bPase is inhibited. Activity of the<br />
kinase responsible for the phosphorylation is itself controlled by cAMP in aglucosemediated<br />
regulatory cascade similar to that shown in Fig. 12.14 on page 437. In S.<br />
cerevisiae, avery close correlation exists between the intracellular concentration of<br />
F2,6bP <strong>and</strong> the production of ethanol. This suggests that F2,6bP is of predominant<br />
importance in controlling glycolytic flux.<br />
12.5.2 Hexose monophosphate (pentose phosphate) pathway<br />
The hexose monophosphate pathway is an alternative route to glycolysis for sugar<br />
metabolism (Fig. 12.7). The pathway is often referred to as ashunt since it diverts a<br />
proportion of glucose from the main glycolytic path <strong>and</strong> returns metabolites at the level<br />
of triose phosphate <strong>and</strong> fructose 6-phosphate. The first part of the pathway is<br />
irreversible <strong>and</strong> catalyses the oxidation of glucose 6-phosphate into a pentose<br />
phosphate. The oxidation generates reducing power in the form of NADPH+H + . This<br />
is utilized in anabolic reactions, which have a specific requirement for this pyridine<br />
nucleotide.<br />
The second part of the pathway involves a series of mostly reversible interconversions<br />
of pentose phosphates, hexose phosphates <strong>and</strong> triose phosphates. This provides a route for<br />
the assimilation of pentoses, in those yeast strains (not S. cerevisiae) capable of utilizing<br />
them. In addition, it provides precursors for the biosynthesis of some vitamins, purine <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glucose<br />
CH2O P<br />
H H H<br />
HO OH H OH<br />
H OH<br />
Glucose<br />
6-phosphate<br />
ZWF1<br />
6-Phosphogluconolactone<br />
NADP + NADPH + H +<br />
GLYCOLYSIS Xylulose<br />
5-phosphate<br />
P OCH2 CH2OH O<br />
H HO<br />
H OH<br />
OH H<br />
CH2O P<br />
H H<br />
HO OH<br />
H<br />
H OH<br />
CH 2OH C O HCOH HCOH CH2O P<br />
CHO HCOH CH2O P<br />
Ribuluose 5-phosphate<br />
RPE1<br />
Glyceraldehyde<br />
3-phosphate<br />
Fructose<br />
6-phosphate<br />
Glyceraldehyde<br />
3-phosphate<br />
TKL1<br />
TKL2<br />
6-Phosphogluconate<br />
GND1 NADP<br />
GND2<br />
CO2<br />
+<br />
NADPH + H +<br />
Ribose<br />
5-phosphate<br />
TLK1<br />
TLK2<br />
Sedoheptulose<br />
7-phosphate<br />
TAL1<br />
Erythrose<br />
4-phosphate<br />
Xylulose<br />
5-phosphate<br />
pyrimidine nucleotides as well as the aromatic amino acids, phenylalanine, tryptophan<br />
<strong>and</strong> tyrosine.<br />
The proportion of carbon flow that is diverted through the hexose monophosphate<br />
shunt is dependent on the cellular requirements for anabolism. Since yeast growth during<br />
brewery fermentations is modest, it seems likely that this proportion is commensurately<br />
small. For yeast growing under fully oxidative conditions, where growth yields are<br />
relatively higher, the proportion of carbon diverted through the hexose monophosphate<br />
shunt will be higher. In addition, where the growth medium is relatively simple, anabolic<br />
requirements are increased, as is the need for NADPH. Bruinenberg et al. (1983)<br />
estimated that under aerobic conditions 2±3% of carbon flow had to pass through the<br />
hexose monophosphate shunt to fulfil the anabolic requirements of yeast cells growing on<br />
glucose.<br />
12.5.3 Tricarboxylic acid cycle<br />
In oxidative metabolism, some of the pyruvate derived from glycolysis is oxidized to<br />
acetyl Coenzyme A (acetyl-CoA), a reaction catalysed by the pyruvate dehydrogenase<br />
complex. The acetyl units are then completely oxidized to two molecules of carbon<br />
dioxide in a series of reactions variously termed the citric acid cycle, tricarboxylic acid<br />
cycle (TCA) or Krebs cycle. Coenzyme A, serves as a carrier of acyl groups in many<br />
enzymatic reactions. It contains the growth factor pantothenic acid. The latter possesses a<br />
terminal thiol group, which is capable of forming thioesters with acyl groups. Pyruvate<br />
O<br />
H2O<br />
–<br />
CH2O P CHOH HCOH HOCH HCOH COO<br />
CH 2OH C O HCOH HCOH CH2O P<br />
RKI1<br />
CHO HCOH HCOH HCOH CH 2O P<br />
CH 2OH C O HCOH (HCOH)3 CH2O P<br />
CHO HCOH HCOH CH 2O P<br />
Fig. 12.7 The hexose monophosphate pathway (pentose phosphate pathway). The genes <strong>and</strong><br />
enzymes responsible for each step are: ZWF1, glucose 6-phosphate dehydrogenase; GND1, GND2,<br />
6-phosphogluconate dehydrogenase; RPE1, D-ribulose 5-phosphate; RKI1, D-ribose 5-phosphate<br />
ketoisomerase; TKI1, TKI2, transketolase; TAL1, transaldolase.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
dehydrogenase is a multi-enzyme complex, which is located in the mitochondrial matrix.<br />
It shares structural <strong>and</strong> functional similarities with glycine dehydrogenase, -<br />
ketoglutarate dehydrogenase <strong>and</strong> branched-chain -ketoacid dehydrogenase. All contain<br />
a common component, lipoamide dehyrogenase, which contains the reducible prosthetic<br />
group, flavin adenine dinucleotide (FAD).<br />
The conversion of pyruvate to acetyl-CoA is an oxidative decarboxylation reaction.<br />
Following the initial release of CO2, the coenzyme, thiamine pyrophosphate (TPP,<br />
vitamin B1) acts as a carrier of the resultant -hydroxyethyl group. The latter is reduced<br />
<strong>and</strong> the bound acetyl group is transferred to another coenzyme, lipoic acid, in a reaction<br />
catalysed by lipoate acetyl transferase. The acetyl group is then transferred to coenzyme<br />
A <strong>and</strong> acetyl CoA is released. The reduced lipoic acid coenzyme is re-oxidized by<br />
lipoamide dehydrogenase. Finally, the reduced FAD prosthetic group of the latter enzyme<br />
is re-oxidized by NAD + <strong>and</strong> NADH+H + is liberated.<br />
Acetyl-CoA enters the tricarboxylic acid cycle (TCA) in a reaction in which it<br />
condenses with oxaloacetate to form citrate (Fig. 12.8). For each turn of the cycle, two<br />
–<br />
OOC CH<br />
HC COO –<br />
CH 2COO –<br />
CH 2COO –<br />
Pyruvate<br />
CO2 + H2O<br />
NAD<br />
CoASH<br />
Pyruvate carboxylase ATP<br />
CO2<br />
+<br />
NADPH + H +<br />
ADP + Pi<br />
NADPH + H +<br />
NAD +<br />
Malate<br />
H2O<br />
Fumarate<br />
SDH1<br />
FADH<br />
SDH2<br />
FAD SDH3<br />
SDH4<br />
Succinate<br />
CoASH<br />
CH 2COO –<br />
HO CH<br />
GTP<br />
CH 2<br />
CH 2CO S CoA<br />
COO –<br />
COO –<br />
GDP<br />
Succinyl-CoA CO2<br />
NADPH + H +<br />
NAD +<br />
FUM1<br />
COO –<br />
COO –<br />
CO<br />
CH2 Oxaloacetate<br />
MDH1<br />
MDH2<br />
MDH3<br />
CoASH<br />
LPD1<br />
KGD1<br />
KGD2<br />
α-Ketoglutarate<br />
CH 2COO –<br />
CH2 C O COO –<br />
Acetyl-CoA<br />
CIT1<br />
CIT2<br />
CIT3<br />
CoASH<br />
Citrate<br />
cis-aconitate<br />
H2O<br />
H2O<br />
Isocitrate<br />
NAD(P)<br />
Oxalo-succinate<br />
+<br />
CH<br />
IDH1<br />
IDH2<br />
HOC H<br />
IDP1<br />
IDP2<br />
NAD(P)H<br />
CO2<br />
CH 3<br />
CO<br />
CH 3 CO S CoA<br />
ACO1<br />
COO –<br />
CH 2COO –<br />
HO C<br />
CH2COO –<br />
COO –<br />
CH2COO CH<br />
–<br />
COO –<br />
C O COO –<br />
LPD1<br />
PDB1<br />
LAT1<br />
CH 2COO –<br />
C COO –<br />
OH<br />
CH 2COO –<br />
Fig. 12.8 The tricarboxylic acid (TCA) cycle. The genes <strong>and</strong> enzymes responsible for each step<br />
are: LPD1, PDA1, PDB1, LAT1, component parts of pyruvate dehydrogenase complex; CIT1,<br />
CIT2, CIT3, citrate synthase; ACO1, aconitase; IDH1, IDH2, NAD-dependent isocitrate<br />
dehydrogenase; IDP1, IDP2, NADP-dependent isocitrate dehydrogenase; KGD1, KGD2, LPD1,<br />
component enzymes of -ketoglutarate dehydrogenase (LPD1 lipoamide dehydrogenase); SDH1,<br />
SDH2, SDH3, SDH4, succinate dehydrogenase; FUM1, fumarase; MDH1, MDH2, MDH3, malate<br />
dehydrogenase.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
COO –<br />
COO –<br />
COO –
molecules of CO2 are liberated <strong>and</strong> reducing power is generated in the form of<br />
NADH+H + <strong>and</strong> FADH. Aproportion of the reduced coenzymes is re-oxidized by the<br />
electron transport chain. The passage of electrons/H + down the chain to oxygen, where it<br />
is oxidized to water, is aprocess that generates large quantities of ATP (Section 12.5.4).<br />
Several of the steps in the TCA cycle are catalysed by isozymes, multiple forms of the<br />
same enzyme, each encoded by adistinct gene. This apparent redundancy is explainable<br />
in that, likeglycolysis, the TCA cycle is anamphibolicpathway that serves both anabolic<br />
<strong>and</strong> catabolic functions. Typically, individual isozymes fulfil different roles in cellular<br />
metabolism. Thus, some function in the oxidative cycle <strong>and</strong> generate reducing power for<br />
the electron transport chain. Others are part of another pathway, the glyoxylate cycle,<br />
which has arole in gluconeogenesis (Section 12.5.6).<br />
SomeTCA cycleisozymeshavebiosynthetic roles<strong>and</strong>provide precursorsused forthe<br />
formation of amino acids such as aspartate <strong>and</strong> glutamate. Other TCA cycle enzymes<br />
contribute to anaplerotic pathways. These are metabolic sequences which literally have a<br />
`filling up' function. Thus, some of the carbon flux through the TCA cycle is utilized in<br />
biosynthetic reactions. Forthe oxidative cycle to continue it isessential to replenish these<br />
intermediates via alternative routes. Commonly, the different isozymes occupy distinct<br />
intracellular compartments or organelles. These organelles have particular metabolic<br />
pathways associated with them <strong>and</strong> these underpin the functions of the organelles. The<br />
oxidative TCA cycle is located within mitochondria <strong>and</strong> glyoxylate cycle enzymes are<br />
found within peroxisomes. Enzymes associated with biosynthesis <strong>and</strong> anaplerosis are<br />
usually located in the cytosol.<br />
Citrate synthase is the rate-limiting step in the TCA cycle. The enzyme, encoded by<br />
the CIT1 gene, is mitochondrial <strong>and</strong> part of the oxidative TCA cycle. The CIT2 gene<br />
product is peroxisomal <strong>and</strong> part of the glyoxylate cycle. The CIT1 citrate synthase is<br />
subject to glucose repression. This phenomenon is exhibited by many of the other<br />
mitochondrial isozymes, for example, aconitase <strong>and</strong> malate dehydrogenase (MDH1 <strong>and</strong><br />
MDH2). In addition, MDH2 malate dehydrogenase is subject to catabolite inactivation,<br />
whereas, -ketoglutarate dehydrogenase activity is regulated by catabolite repression.<br />
Four isozymes of isocitrate dehydrogenase are synthesized by yeast, two<br />
mitochondrial <strong>and</strong> two cytosolic. One of each pair is specific for NAD + whereas, the<br />
other two are specific for NADP + .Only the NAD + -linked enzymes are active in the<br />
oxidative TCA cycle. The NADP + -linked enzymes probably have anaplerotic roles, as<br />
does the MDH2 malate dehydrogenase. The peroxisomal MDH3 enzyme appears to be<br />
involved in the oxidation of NADH, which is produced from -oxidation of fatty acids<br />
(Section 12.7.1).<br />
The supply of oxaloacetate for the oxidative TCA cycle (Fig. 12.8) is ensured<br />
primarily by the action of the anaplerotic enzyme, pyruvate carboxylase. This catalyses<br />
the condensation of pyruvate <strong>and</strong> CO2 to form oxaloacetate. The enzyme contains biotin<br />
as a prosthetic group. Biotin serves as a carrier of carboxyl groups. Energy is provided by<br />
the breakdown of a molecule of ATP. Pyruvate carboxylase is inhibited by aspartate <strong>and</strong><br />
allosterically activated by acetyl-CoA <strong>and</strong> ATP. The inhibition is an end-product feedback<br />
control mechanism since aspartate is derived from the transamination of<br />
oxaloacetate. Stimulation by acetyl-CoA <strong>and</strong> ATP ensures that there is always a<br />
sufficient supply of oxaloacetate to maintain an adequate supply of substrates for citrate<br />
synthase. Transamination reactions between glutamate <strong>and</strong> pyruvate to yield alanine <strong>and</strong><br />
oxaloacetate may also be of anaplerotic significance.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HO CH<br />
CH2 COO –<br />
COO –<br />
H2O<br />
–<br />
OOC CH<br />
HC COO –<br />
+<br />
NADH + H<br />
Oxaloacetate<br />
NAD<br />
MDH3<br />
Malate<br />
CoASH<br />
Fumarate<br />
2H<br />
MLS1<br />
CH 2<br />
COO –<br />
COO –<br />
Acetyl-CoA<br />
Glyoxylate<br />
Succinate<br />
Acetate<br />
Ethanol<br />
Fatty acids<br />
12.5.4 Electron transport <strong>and</strong> oxidative phosphorylation<br />
In the terminal stage of the oxidative catabolism of sugars, the reduced redox coenzymes,<br />
NADH+H + <strong>and</strong> FADH (reduced FAD; 4.94), arising from the TCA cycle <strong>and</strong> glycolysis,<br />
are re-oxidized. The process is mediated by a series of redox carriers <strong>and</strong> it culminates in<br />
the reduction of molecular oxygen to water. Together, the redox carriers constitute the<br />
respiratory or electron transport chain (Fig. 12.9). Consecutive components of the<br />
electron transport chain have progressively more positive st<strong>and</strong>ard redox potentials,<br />
which facilitates the ordered transfer of electrons. Transfer of electrons down the<br />
transport chain generates energy, a proportion of which is retained in the form of the<br />
high-energy bonds of ATP. The process of energy transduction is termed oxidative<br />
phosphorylation. It can be summarised in the following equation:<br />
DH2 ‡ 1<br />
2 O2 ‡ nADP ! D ‡ nATP ‡ H2O<br />
CO<br />
CHO<br />
COO –<br />
CH 2COO –<br />
CH 2COO –<br />
Citrate<br />
Isocitrate<br />
DH2 is a hydrogen donor. The value of n is a variable <strong>and</strong> is dependent on the tightness of<br />
coupling between respiration <strong>and</strong> phosphorylation <strong>and</strong> the nature of the donor. The<br />
efficiency of phosphorylation is commonly expressed as the P:O ratio, which is the<br />
number of ATP molecules generated per oxygen atom utilized.<br />
The redox carriers are a diverse group of compounds that share the common property<br />
of having a reversibly reducible component. Cytochromes are haemoproteins in which<br />
the prosthetic group, haem, is a tetracyclic pyrrole, containing an atom of iron, which can<br />
be reversibly reduced from the ferric to the ferrous form. Ubiquinone (Coenzyme Q) is a<br />
hydrophobic quinone, which can be reversibly reduced to the quinol form. Iron sulphur<br />
proteins also undergo transitions between the ferrous <strong>and</strong> ferric states. Flavoproteins<br />
ICL1<br />
COO –<br />
CH2COO –<br />
HO C<br />
CH2COO –<br />
CH 2COO –<br />
CH<br />
HOC H<br />
Fig. 12.9 The glyoxylate cycle. The genes <strong>and</strong> enzymes are: ICL1, isocitrate lyase; MLS1, malate<br />
synthase; MDH3, malate dehydrogenase (cytosolic). Other stages are catalysed by enzymes of the<br />
tricarboxylic acid cycle.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
COO –<br />
COO –
contain prosthetic groups flavin mononucleotide (FMN) or flavin adenine dinucleotide<br />
(FAD). Both contain areversibly reducible isoalloxazine group within the riboflavin<br />
moiety.<br />
The electron transport chain consists of five complexes, which are located within the<br />
inner mitochondrial membrane (Fig. 12.9). The first complex, NADH-CoQ reductase<br />
accepts electrons from NADH, generated by the mitochondrial TCA cycle. The electrons<br />
are passed on to apool of ubiquinone causing the latter to be reduced to ubiquinol. The<br />
ubiquinone pool also accepts electrons from the second complex, succinate dehydrogenase,<br />
located on the inner surface of the inner mitochondrial membrane. In yeast, this<br />
dehydrogenase also shows activity towards -glycerophosphate. Ubiquinols are reoxidized<br />
by transfer of electrons to the third complex, CoQ-cytochrome Creductase. In<br />
yeast, this complex contains cytochromes b<strong>and</strong> c 1<strong>and</strong> an iron-sulphur protein. The<br />
cytochrome cpool mediates transferofelectrons betweenthe third<strong>and</strong> fourth complexes.<br />
The latter is cytochrome coxidase, which contains cytochromes a, a3 <strong>and</strong> acopper<br />
metalloprotein. Cytochrome oxidase completes the process by transferring electrons to<br />
oxygen, generating water.<br />
The fifth complex is an ATP synthase, the activity of which is coupled to the energy<br />
liberated by the controlled flow of electrons. This process is accomplished according to<br />
the principles of the chemiosmotic theory (Mitchell, 1979). This holds that the electron<br />
chain complexes are arranged spatially within the mitochondrial inner membrane such<br />
that as electrons flow down their electrical potential, protons are translocated from the<br />
inside to the outside of the membrane. Since the membrane is relatively impermeable to<br />
protons <strong>and</strong> other charged species, the electrogenic pumping of protons generates a<br />
transmembraneelectrochemicalpotentialdifference.Thishasbothelectrical(charge)<strong>and</strong><br />
chemical (proton concentration) components. This thermodynamic potential drives the<br />
synthesis of ATP via areversible proton-translocating ATPase or ATP synthase.<br />
Complexes 1, 3<strong>and</strong> 4are associated with proton pumping <strong>and</strong> hence, indirectly with<br />
ATP generation, termed sites I, II <strong>and</strong> III, respectively. Theoretically, each pair of<br />
electrons traversing the whole of the respiratory chain could generate three molecules of<br />
ATP. Electrons arising in the mitochondrial matrix from the oxidation of succinate<br />
bypass the first phosphorylation site. In <strong>practice</strong>, actual yields of ATP are lower. In some<br />
yeast genera, for example, C<strong>and</strong>ida utilis, phosphorylation site Iis present during growth<br />
under certain conditions of nutrient limitation. However, in S. cerevisiae, including<br />
brewing strains, the existence of site I has been questioned (Guerin, 1991). Unlike<br />
mammalian cells, mitochondria of S. cerevisiae oxidize exogenous NAD(P)H, directly<br />
via NAD(P) + dehyrogenases located in the outer surface of the inner membrane. These<br />
deliver electrons to the common ubiquinone pool <strong>and</strong> therefore, bypass site I<br />
phosphorylation. Systems for transporting NADH into mitochondria do exist in S.<br />
cerevisiae, for example, the malate ± aspartate shuttle (Fig. 12.10). This utilizes a<br />
combination of malate dehydrogenase <strong>and</strong> transamination reactions to transfer reducing<br />
equivalents from the cytosol to the mitochondria. In S. cerevisiae, it serves to control the<br />
concentration of NADH in the cytosol.<br />
The individual complexes of the respiratory chain are susceptible to inhibition by a<br />
variety of compounds. These have been used as tools to identify which components are<br />
present <strong>and</strong> their order within the respiratory chain. For example, rotenone inhibits<br />
complex I. The lack of effect of this compound on oxidative phosphorylation in S.<br />
cerevisiae is primary evidence for suggesting that complex I is absent in this yeast.<br />
Cyanide, azide <strong>and</strong> antimycin A, inhibit complexes III, <strong>and</strong> IV. Oligomycin inhibits<br />
complex V, ATP synthase.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
NADH + H +<br />
NAD +<br />
CO<br />
Cytosol Mitochondrial<br />
membrane<br />
Inner<br />
CH 2<br />
COO –<br />
COO –<br />
Oxaloacetate<br />
α-Ketoglutarate<br />
CH 2COO –<br />
CH 2<br />
C O COO –<br />
Glutamate<br />
Aspartate<br />
1<br />
Glutamate<br />
Aspartate<br />
2<br />
Malate Malate<br />
HC CH COO –<br />
CH 2<br />
COO –<br />
H 2N<br />
CH COO –<br />
CH 2<br />
H 2N CO COO –<br />
CH 2<br />
COO –<br />
Oxaloacetate<br />
α-Ketoglutarate<br />
1. Glutamate/Aspartate carrier<br />
2. Dicarboxylic acid carrier<br />
NADH + H +<br />
Fig. 12.10 The malate-aspartate shuttle system for transferring reducing equivalents from the cytosol to mitochondria.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
CH 2<br />
COO –<br />
NAD +
In many organisms alternative respiratory systems can be detected that are resistant to<br />
cyanide inhibition. There are two types of cyanide insensitive respiration, which are<br />
differentiated based on susceptibility to inhibition by salicylylhydroxamic acid (SHAM).<br />
The SHAM-sensitive pathway has been detected in several yeast strains, including<br />
Yarrowia lipolytica <strong>and</strong> stationary phase cultures of many strains, including C<strong>and</strong>ida<br />
utilis. The SHAM insensitive pathway has been detected in S. cerevisiae, S. pombe,<br />
Kluyveromyces lactis, Hansenula saturnus <strong>and</strong> Endomycopsis capsularis. As would be<br />
predictedfromthelackofsensitivitytocyanidetheserespiratorypathwaysarenotcoupled<br />
to energy transduction. It is possible that they function as part of cellular redox control.<br />
12.5.5 Fermentative sugar catabolism<br />
During growth of S. cerevisiae on glucose <strong>and</strong> other repressing carbohydrates, the major<br />
product of sugar catabolism is ethanol. Aproportion of pyruvate derived from glycolysis<br />
is decarboxylated to acetaldehyde, a reaction catalysed by pyruvate decarboxylase.<br />
Acetaldehyde is then reduced to ethanol in areaction performed by NAD + -linked alcohol<br />
dehydrogenase (Fig. 12.11). In this mode of metabolism ATP is derived solely from<br />
glycolytic substrate level phosphorylation <strong>and</strong> alcohol dehydrogenase is the major redox<br />
control route for regeneration of NAD + .<br />
In S. cerevisiae growing on glucose <strong>and</strong> to alesser extent other fermentative sugars,<br />
fermentative metabolism is predominant irrespective of the presence of oxygen. Under<br />
these conditions, the phenomenon of glucose repression ensures that the genes encoding<br />
purely oxidative pathways such as respiratory oxidative phosphorylating electron<br />
transport chain are not expressed (Section 12.5.8). Furthermore, enzymes responsible<br />
for the utilization of oxidative carbon sources such as ethanol <strong>and</strong> glycerol are not<br />
synthesized <strong>and</strong> mitochondrial development is arrested. If the glucose concentration falls<br />
to a low level (< 0.2% w/w) the metabolism shifts <strong>and</strong> becomes `derepressed'. In this<br />
PDA1<br />
PDHβ1<br />
LAT1<br />
Acetyl-CoA<br />
CH 3 CO S CoA<br />
PDX1<br />
LPD1<br />
NAD +<br />
NADH + H +<br />
CoASH<br />
CoASH<br />
ASC2<br />
ADP ATP<br />
Acetate<br />
CH 3<br />
COO –<br />
CH 3<br />
Pyruvate<br />
CO2<br />
CO<br />
COO –<br />
ALD6<br />
PDC1<br />
PDC5<br />
PDC6<br />
NAD(P)H + H + NAD(P) +<br />
Acetaldehyde<br />
ADH1<br />
ADH2<br />
ADH3<br />
ADH4<br />
CO2<br />
Ethanol<br />
CH 3CHO<br />
NADH + H +<br />
NAD +<br />
Fig. 12.11 Enzymes of pyruvate catabolism. The genes <strong>and</strong> enzymes for each step are: PDA1,<br />
PDH 1, LAT1, PDX1, LPD1, pyruvate dehydrogenase complex; PDC1, PDC5, PDC6, pyruvate<br />
decarboxylase; ADH1, ADH2, ADH3, ADH, 4, alcohol dehydrogenase; ALD6, aldehyde<br />
dehydrogenase (cytosolic); ACS2, acetyl-CoA synthetase (cytosolic).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
case, the effects of the glucose signal are abolished <strong>and</strong> induction of the respiratory<br />
metabolic machinery occurs. In the presence of oxygen, the utilization of oxidative<br />
substrates, such as ethanol, is coupled to phosphorylation by the electron transport chain.<br />
The change in metabolism from fermentative <strong>and</strong> glucose-consuming to oxidative <strong>and</strong><br />
ethanol-consuming is an example of diauxie, or adiauxic shift. In other words, the<br />
phenomenon of biphasic growth where after the exhaustion of an initial substrate there is<br />
alag phase during which phenotypic adaptation occurs. This results in asecond period of<br />
growth during which an alternative substrate is utilized.<br />
The branch-point between oxidative <strong>and</strong> fermentative sugar catabolism occurs at the<br />
level of pyruvate. Pyruvate decarboxylase is encoded by three genes, PDC1, PDC5 <strong>and</strong><br />
PDC6. Of these, PDC1 is the predominant isozyme <strong>and</strong> is strongly expressed in the<br />
presence of glucose. PDC5 appears to be fully expressed only if PDC1 is impaired. PDC<br />
6 expression occurs only during growth on ethanol. Like pyruvate dehydrogenase,<br />
pyruvate kinase requires the cofactor, thiamine pyrophosphate for activity.<br />
Yeast alcohol dehydrogenase is encoded by four genes. The ADH1 product is<br />
associatedwith thereductionofacetaldehydetoethanol sinceitisinducedstronglyinthe<br />
presence of glucose. ADH2 alcohol dehydrogenase is implicated in ethanol utilization<br />
since it is repressed by glucose <strong>and</strong> only active in derepressed cells. The roles of ADH3<br />
<strong>and</strong> ADH4 are not clear. The former is amitochondrial enzyme which is induced by<br />
glucose, although less strongly than ADH1. The ADH4 enzyme requires Zn 2+ for<br />
activity.Itcanusuallybedetected inbrewingstrainsofS.cerevisiae butnot inlaboratory<br />
strains. It is possible that in brewery fermentations these enzymes are involved in the<br />
formation of higher alcohols <strong>and</strong>/or the reduction of vicinal diketones (Section 12.10).<br />
Possibly, these enzymes have a role in redox control during anaerobiosis.<br />
Under fermentative conditions, the cell has a requirement for acetyl-CoA for<br />
biosynthetic reactions, in particular, syntheses of amino acids <strong>and</strong> fatty acids. Albeit, the<br />
total requirement for acetyl-CoA is reduced since the respiratory pathways are not<br />
operational. Nevertheless, there must be metabolic regulation of carbon flow between<br />
ethanol formation via pyruvate decarboxylase <strong>and</strong> pathways leading to the formation of<br />
acetyl-CoA. In yeast, this may be achieved by regulation of the activities of the pyruvate<br />
dehydrogenase complex <strong>and</strong> pyruvate decarboxylase. In addition, a bypass mechanism<br />
may be implicated, whereby acetyl-CoA is derived from acetaldehyde via the concerted<br />
action of acetaldehyde dehydrogenase <strong>and</strong> acetyl-CoA synthetase (Figure 12.11).<br />
At low glucose concentrations there is a concomitant low concentration of pyruvate.<br />
Consequently, the pyruvate dehydrogenase route predominates since this enzyme has a<br />
higher affinity for pyruvate than pyruvate decarboxylase (Postma et al., 1989). At higher<br />
glucose concentrations, pyruvate concentrations also increase <strong>and</strong> an increasing<br />
proportion of the carbon flow is directed towards the formation of acetaldehyde, via<br />
pyruvate decarboxylase. This is then converted to acetate <strong>and</strong> acetyl-CoA through the<br />
bypass route. At still higher glucose concentrations, the acetyl-CoA synthetase reaction<br />
becomes rate-limiting <strong>and</strong> ethanol formation occurs, via alcohol dehydrogenase. This<br />
shift is further favoured by the increase in pyruvate decarboxylase activity due to glucose<br />
activation.<br />
Mammalian pyruvate dehydrogenases are subject to regulation via reversible<br />
phosphorylation. A similar mechanism in yeast has not been demonstrated, conclusively.<br />
Some control on pyruvate dehydrogenase activity may be exerted at the level of<br />
transcription. Under fermentative conditions, pyruvate dehydrogenase may be subject to<br />
limited transcription. The latter is controlled in concert with the enzymes of amino acid<br />
biosynthesis.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Pyruvate decarboxylase is acytosolic enzyme, whereas, pyruvate dehydrogenase is<br />
mitochondrial, as are the enzymes of amino acid biosynthesis which utilize pyruvate <strong>and</strong><br />
acetyl-CoA. For the acetaldehyde bypass to be effective acetyl-CoA must cross the<br />
mitochondrial membrane. Mitochondrial membranes are accessible to acetate but not<br />
acetyl-CoA. Acetyl-CoA synthetases <strong>and</strong> deacylases occur in both the cytosolic <strong>and</strong><br />
mitochondrial compartments. In some yeast strains, acetyl units may be transported into<br />
mitochondria in the form of acetylcarnitine.<br />
12.5.6 Gluconeogenesis <strong>and</strong> the glyoxylate cycle<br />
Under aerobic conditions, yeast can utilize anumber of non-sugar sources, such as<br />
ethanol, glycerol <strong>and</strong> lactate as sole sources of carbon. For growth to proceed under these<br />
conditions the cell must synthesize glycolytic intermediates, some of which are<br />
precursors ofessential anabolicmetabolites.This requires glycolysis tooperateinreverse<br />
<strong>and</strong> this process is termed gluconeogenesis (Gancedo <strong>and</strong> Gancedo, 1997). The majority<br />
of the enzymes of glycolysis are freely reversible. However, phosphofructokinase <strong>and</strong><br />
pyruvate kinase are not. These non-reversible steps are bypassed by two specific<br />
gluconeogenic enzymes. Fructose 1,6-bisphosphatase catalyses the conversion of<br />
fructose 1,6-bisphosphate to fructose 6-phosphate <strong>and</strong> phospho-enol-pyruvate carboxykinase<br />
catalyses the conversion of oxaloacetate to phospho-enol-pyruvate (Fig. 12.5).<br />
Glycolysis <strong>and</strong> gluconeogenesis do not occur simultaneously <strong>and</strong> therefore the<br />
direction of carbon flow is regulated. As befits such acrucial crossroad in metabolism,<br />
control mechanisms are many <strong>and</strong> stringent. The presence of exogenous glucose at very<br />
low concentration (0.2mM) totally abolishes gluconeogenesis. Under these conditions,<br />
the transcription ofFBP1 (fructose 1,6-bisphosphatase) <strong>and</strong>PCK1 (phosphoenolpyruvate<br />
carboxykinase) is repressed. The glucose signal does not involve the Ras, cyclic AMP<br />
system. Phosphorylation of glucose is required since the repression of the gluconeogenic<br />
enzymes is abolished if hexokinase 1, 2<strong>and</strong> glucokinase are absent.<br />
When glucose is added to yeast growing gluconeogenically on an oxidative substrate<br />
fructose 1,6-bisphosphatase <strong>and</strong> phosphoenolpyruvate carboxykinase are degraded by<br />
proteases, aphenomenon termed catabolite inactivation. Both enzymes are reversibly<br />
inactivated by phosphorylation. Fructose 1,6bisphosphatase is strongly inhibited by<br />
AMP <strong>and</strong> fructose 2,6bisphosphate. The latter metabolite is apositive effector of 6phosphofructo-1-kinase.<br />
The regulation of the enzymes which produce <strong>and</strong> degrade<br />
fructose 2,6-bisphosphate are part of a glucose- mediated, cyclic AMP-dependent<br />
reversible phosphorylation cascade, as described in Section 12.5.8.<br />
The glyoxylate cycle is apart of the gluconeogenic pathway, which is essential for<br />
growthontwo-carbonsubstratessuchasethanol<strong>and</strong>acetate.Essentially,thepathwayisa<br />
short cut through the TCA cycle that bypasses those irreversible steps that result in aloss<br />
of carbon as CO2. It involves two specific glyoxylate cycle enzymes, isocitrate lyase <strong>and</strong><br />
malate synthase (Fig. 12.12). Both enzymes are subject to catabolite inactivation when<br />
glucose is added to yeast growing gluconeogenically. Isocitrate lyase <strong>and</strong> the MDH3<br />
malate dehydrogenase are both induced by ethanol <strong>and</strong> repressed by glucose. The<br />
glyoxylate cycle is not operative in brewing fermentations.<br />
12.5.7 Storage carbohydrates<br />
Under some conditions, the carbohydrates trehalose <strong>and</strong> glycogen accumulate in yeast <strong>and</strong><br />
under other conditions, they are degraded (Lillie <strong>and</strong> Pringle, 1980). Since they do not<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ADP + Pi<br />
H +<br />
Mitochondrial<br />
ATPase<br />
(Complex 5)<br />
H +<br />
ATP<br />
H +<br />
Site I<br />
NADH-CoQ<br />
reductase<br />
(Complex 1)<br />
Mitochondrial<br />
NADH<br />
NAD(P)H + H +<br />
NAD(P) +<br />
NAD(P)H<br />
dehydrogenases<br />
Ubiquinone<br />
pool<br />
Succinate<br />
dehydrogenase<br />
(Complex 2)<br />
Succinate Malate<br />
TCA cycle<br />
H +<br />
Site II Site III<br />
H +<br />
CoQ-cytochrome C<br />
reductase<br />
(Complex 3)<br />
Outer mitochondrial<br />
membrane<br />
Cytochrome C<br />
pool<br />
Inner mitochondrial<br />
membrane<br />
Mitochondrial<br />
matrix<br />
O2<br />
H2O<br />
Cytochrome C<br />
oxidase<br />
(Complex 4)<br />
Fig. 12.12 Representation of electron transfer <strong>and</strong> oxidative phosphorylation in mitochondria (adapted from Alex<strong>and</strong>er <strong>and</strong> Jefferies, 1990). Note the presence of<br />
the three proton-pumping sites <strong>and</strong> the reverse flow of protons that permits ATP formation.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glucose<br />
1-phosphate<br />
UTP<br />
Glucose<br />
Glucose 6-phosphate Trehalose 6-phosphate<br />
PGM1<br />
PGM2<br />
UGP1<br />
UDP-glucose<br />
GSY1<br />
PPi GSY2<br />
Glycogen<br />
(n–1)<br />
TPS1<br />
have structural roles, it has been assumed that both function as reserve materials. This<br />
appears to be an over-simplistic interpretation. Glycogen is a polymer of -D-glucose with<br />
a molecular weight of approximately 10 8 . It consists of chains of 10 14 residues of -Dglucose<br />
joined by 1 ! 4 linkages. These chains are cross-linked by (1 ! 6)- -Dglucosidic<br />
linkages <strong>and</strong> therefore there are structural similarities to amylopectin. In<br />
brewing strains of S. cerevisiae, during fermentation, up to 4% of wort sugars are<br />
converted to glycogen <strong>and</strong> it can account for up to 20 30% of the dry weight of the cell<br />
(Quain <strong>and</strong> Tubb, 1982). There are two pools of glycogen in S. cerevisiae. The first is<br />
soluble <strong>and</strong> its concentration is modulated in response to changes in physiological state.<br />
The second is only solubilized by treatment with acid. It has a structural role, being<br />
covalently linked to cell wall -glucans (Arvindekar <strong>and</strong> Patil, 2002). Trehalose ( -Dglucopyranosyl-1,<br />
1- -D-gluconopyranoside) is a disaccharide consisting of two molecules<br />
of D-glucose. In commercial preparations of bakers' yeast, trehalose accounts for<br />
15 20% of the cell dry weight. In brewing yeast after fermentation, trehalose<br />
concentrations are usually quite modest, typically 2 3% of the dry weight. In yeast<br />
recovered from very high-gravity worts (25 ëPlato, SG c. 1106.1), trehalose levels of<br />
20 25% of the cell dry weight have been reported (Majara et al., 1996a, b).<br />
Glycogen <strong>and</strong> trehalose are both synthesized from glucose 6-phosphate (Fig. 12.13).<br />
Both biosynthetic pathways utilize uridine triphosphate (UTP) to generate the glucosedonor<br />
molecule, uridine diphosphate-glucose (UDP-glucose). This reaction is catalysed<br />
by UDP-glucose pyrophosphorylase. Trehalose is synthesized by the action of two<br />
enzymes, trehalose 6-phosphate synthase (TPS1) <strong>and</strong> trehalose 6-phosphatase (TPS2).<br />
Degradation is via the hydrolase, trehalase, two forms of which occur in yeast. The<br />
neutral form (NTH1) is cytosolic <strong>and</strong> acidic trehalase (ATH1) is located in vacuoles. A<br />
second <strong>and</strong> possibly minor pathway for trehalose has been reported, which utilizes<br />
H2O<br />
Pi<br />
UDP UTP<br />
GLC3<br />
ATP ADP<br />
Pi<br />
Pi<br />
TPS2<br />
NTH1<br />
ATH1<br />
Trehalose<br />
CH2OH HO<br />
H<br />
H<br />
OH<br />
HO<br />
O<br />
H H H<br />
OH<br />
O HOH2C O OH<br />
H OH<br />
GPH1<br />
Glycogen Glucose 1-phosphate<br />
Fig. 12.13 Synthesis <strong>and</strong> degradation of glycogen <strong>and</strong> trehalose from <strong>and</strong> to glucose. The genes<br />
<strong>and</strong> enzymes are: PGM1, PGM2, phosphoglucomutase; UGP1, UDP-glucose pyrophosphorylase;<br />
GSY1, GSY2, glycogen synthase; GLC3, branching enzyme; GPH1, glycogen phosphorylase;<br />
NTH1, neutral trehalase, ATH1, acidic trehalase (adapted from Francois et al., 1997).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
adenosine diphosphoglucose-dependent trehalose synthase. This is expressed only when<br />
maltose or galactose is the carbon source.<br />
Trehalose synthase <strong>and</strong> phosphatase occur together in a complex. Their gene<br />
promoters contain sequences that are found in other genes subject to expression by heat<br />
shock (Section 12.9). The vacuolar trehalase has an acidic pH optimum <strong>and</strong> is<br />
constitutive. The cytosolic neutral trehalase is activated by reversible phosphorylation in<br />
response to cyclic AMP-dependent protein kinase.<br />
Glycogen biosynthesis begins with areaction catalysed by an initiator protein <strong>and</strong><br />
results in the formation of an -1 !4glycosyl primer molecule from UDPG. The<br />
initiator protein is produced by two genes, CLG1 <strong>and</strong> CLG2, which are the equivalent of<br />
mammalian glycogenin, responsible for de novo glycogen synthesis. The primer is then<br />
elongated with the formation of -1,4linkages by glycogen synthase. The -1 !6<br />
cross-linkages are synthesized by glycogen branching enzyme, a transglycoylase.<br />
Glycogen degradation to glucose 1-phosphate <strong>and</strong> glucose is accomplished by glycogen<br />
debranching enzyme <strong>and</strong> glycogen phosphorylase.<br />
Glycogen synthase activity is regulated by reversible phosphorylation, possibly under<br />
thecontrolofcyclicAMP-dependentkinase,althoughotherkinasesmayalsobeinvolved.<br />
Inthephosphorylatedform,theenzymeislessactive.Dephosphorylationiscatalysedbya<br />
glycogen synthase phosphatase (GLC7). In the active dephosphorylated form, glycogen<br />
synthase is subject to allosteric activation by glucose 6-phosphate. Glycogen synthase<br />
phosphorylase, the glycogen degradative enzyme occurs in aphosphorylated dimeric<br />
active form <strong>and</strong> anon-active dephosphorylated tetrameric form. Cyclic AMP-dependent<br />
kinase may also be involved in the phosphorylation of glycogen synthase phosphorylase,<br />
however, it appears that this operates via asecond specific glycogen phosphorylasedependent<br />
kinase. The latter may also be active towards glycogen synthase.<br />
Accumulationofglycogen<strong>and</strong>trehaloseoccurswhengrowthisrestricted.Inaglucoselimitingmedium,thisoccurswhenapproximatelyhalfthe<br />
glucose hasbeen consumed. In<br />
glucose-rich media, it occurs in the late exponential phase when another nutrient is<br />
limiting<strong>and</strong>duringtheonsetofdiauxie(Croweetal.,1984).Accumulationofglycogenis<br />
precededbyinductionoftheenzymesresponsibleforitsbiosynthesis<strong>and</strong>byactivationof<br />
glycogen synthase by dephosphorylation. These events are accompanied by other global<br />
metabolicchanges,whichitissuggestedindicatethattheaccumulationofglycogenispart<br />
of ageneral nutrient-sensing system (Francois et al., 1997). Thus, the cell responds to<br />
imminent starvation of an essential nutrient by accumulating glycogen <strong>and</strong> by reducing<br />
glycolytic flux, reducing overall protein synthesis <strong>and</strong> inducing heat shock proteins.<br />
The Ras-cyclic AMP cascade system is involved in nutrient sensing. Cyclic AMP<br />
concentration is reduced during growth on glucose. It reaches its lowest level at the onset<br />
of glycogen accumulation indicating that the cyclic AMP-dependent protein kinase<br />
(cPKA) could be implicated. However, glycogen accumulation also requires the activity<br />
of another protein kinase, the Snf1 gene product. This kinase is aglobal regulator whose<br />
activity is essential for the entire phenomenon of derepression. With regard to glycogen<br />
accumulation, the activities of cPKA <strong>and</strong> snf1 are antagonistic. Both of these kinases are<br />
involved in the regulation of the transcription <strong>and</strong> post-translational control of genes,<br />
which include those encoding the enzymes of glycogen synthesis <strong>and</strong> degradation.<br />
Inbrewing,glycogenisusedbyyeasttoprovidemaintenanceenergyduringtheperiod<br />
when yeast is stored in the interval between cropping <strong>and</strong> re-pitching. Glycogen<br />
dissimilation accounts for the small decrease in yeast cell dry weight, which can be seen<br />
inthelatterphaseoffermentationwhengrowthhasceased(Fig.12.1).Inaddition,during<br />
the aerobic phase of fermentation, glycogen reserves are rapidly mobilized. A linear<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
elationship has been demonstrated between the quantities of glycogen utilized <strong>and</strong> sterol<br />
synthesized (Quain <strong>and</strong> Tubb, 1982).<br />
Trehalose concentrations are low in yeast growing exponentially on glucose.<br />
Accumulation occurs during the transition between exponential growth <strong>and</strong> entry into the<br />
stationaryphase.Thesynthetic<strong>and</strong>degradativepathwaysarebothactivesimultaneously<strong>and</strong><br />
the actual concentration at any instant represents the balance between the two. The neutral<br />
trehalaseisactivatedbyphosphorylationbycPKA.Activityofthesyntheticcomplexisalso<br />
regulated by cPKA. This is not by direct phosphorylation of the enzymes <strong>and</strong> it seems that<br />
the cAMP signal pathway regulates transcription of their respective genes.<br />
Trehalose accumulation in yeast occurs in response to heat shock <strong>and</strong> some other<br />
stresses, such as exposure to hydrogen peroxide. It is assumed that this response renders<br />
the cells more resistant to stress. Trehalose confers resistance to elevated temperature in<br />
many species (Lillie <strong>and</strong> Pringle, 1980). This is because trehalose is able to stabilize<br />
membranes, where it prevents phase transition events in lipid bilayers. For maximum<br />
effectiveness, it must be present at both the inner <strong>and</strong> outer surfaces of the membrane<br />
where it binds to polar heads of phospholipids via the sugar hydroxyl groups. S. cerevisiae<br />
possessesatrehalosetransporter.Ithasbeensuggestedthataswellasbeingresponsiblefor<br />
theuptakeofexogenoustrehalose,thecarriermayalsomediateintracellulartransportfrom<br />
the cytosol to the plasma membrane <strong>and</strong> periplasm (Eleutherio et al., 1993).<br />
Following heat shock to S. cerevisiae, trehalose 6-phosphate synthase activity<br />
increases <strong>and</strong> neutral trehalase is inactivated. The elevated biosynthetic activity is due to<br />
increasedexpressionoftheTPS1<strong>and</strong>TPS2genes.Inaddition,thecatalyticactivityofthe<br />
biosynthetic <strong>and</strong> degradative enzymes alters in favour of trehalose accumulation.<br />
12.5.8 Regulation of sugar metabolism<br />
Sugar metabolism is highly regulated. The energy yield from individual catabolic<br />
pathways is very different. During fermentative growth, ATP generation is restricted to<br />
substrate level phosphorylation. For each glucose molecule oxidized via glycolysis, there<br />
is a net gain of two molecules of ATP. In the case of yeast utilizing oxidative<br />
phosphorylation, this value increases to approximately 30 molecules of ATP generated<br />
per molecule of glucose oxidized. Yeasts may be classified on the basis of their preferred<br />
mode of sugar catabolism. Obligate aerobes make exclusive use of the respiratory<br />
pathways <strong>and</strong> are unable to ferment sugars. They include the genera, Rhodotorula,<br />
Lipomyces, Cryptococcus, Rhodosporidium <strong>and</strong> Saccharomycopsis. Facultative anaerobes<br />
may use both respiratory <strong>and</strong> fermentative pathways. This group is further<br />
subdivided based on the proportion of sugars catabolized by each route under aerobic<br />
conditions. Respiratory types are predominant <strong>and</strong> dispose of 70% or more of sugars via<br />
respiration. These include C<strong>and</strong>ida, Hansenula, Kluyveromyces <strong>and</strong> Pichia. Fermentative<br />
yeasts are typified by high rates of sugar metabolism of which 10% or less is catabolized<br />
by respiration. Saccharomyces (including all brewing strains), Brettanomyces <strong>and</strong><br />
Schizosaccharomyces belong to this category.<br />
Anumber of phenotypic effects are recognized which describe the patterns of sugar<br />
catabolism in various genera growing under certain defined conditions. These are<br />
summarized in Table 12.3. The Crabtree effect was originally described in rat ascites<br />
tumour cells where it was observed that the addition of glucose resulted in a reduction in<br />
respiration rate (Crabtree, 1929). Subsequently, this was ascribed to competition for ATP<br />
<strong>and</strong> inorganic phosphate between glycolysis <strong>and</strong> respiration. In the presence of glucose,<br />
cellular requirements for ATP are satisfied by glycolysis. This results in a reduced<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table12.3 Mechanismsfortheregulationofsugarcatabolisminyeast(Boulton<strong>and</strong>Quain,2001)<br />
Mechanism Description<br />
Short-term Crabtree effect Reduced respiration rate in response to glucose pulse<br />
Glucose catabolite repression <strong>and</strong> Suppression of respiration by glucose<br />
inactivation<br />
Pasteur effect Reduction in rate of glycolysis under aerobic conditions<br />
Kluyver effect Obligate aerobic utilization of disaccharides<br />
Custers effect Aerobic stimulation of rate of glucose fermentation<br />
requirement for the translocation into the cytosol of ATP produced in mitochondria by<br />
oxidative phosphorylation. In turn, this restricts the exchange of cytosolic ADP with<br />
mitochondrial ATP. The resultant reduction in mitochondrial ADP concentration restricts<br />
oxidative phosphorylation <strong>and</strong> respiratory rates decline.<br />
This short-term Crabtree effect occurs within minutes of the addition of glucose. A<br />
similar phenomenon occurs in yeast <strong>and</strong> it is accompanied by an immediate increase in<br />
the rate of ethanol production. The mechanism may be as described in the preceding<br />
paragraph, however it seems more likely that the effect resides in the relative affinity for<br />
pyruvate of pyruvate dehydrogenase <strong>and</strong> pyruvate decarboxylase (Section 12.5.5).<br />
The so-called long-term Crabtree effect describes the repression <strong>and</strong> inactivation of<br />
respiratory enzymes in yeast by the presence of glucose. The underlying biochemistry of<br />
this effect is totally different from the short-term Crabtree effect. More properly, it is<br />
described as glucose catabolite repression <strong>and</strong> inactivation. Obligate aerobic yeasts <strong>and</strong><br />
respiratory facultative anaerobic types exhibit little or no glucose repression. These are<br />
termed Crabtree-negative or weakly Crabtree-positive, as appropriate.<br />
S. cerevisiae is aCrabtree-positive yeast. In aerobically growing cultures to which<br />
glucose (>0.2%w/w) has been added, glycolysis becomes the major energy-yielding<br />
pathway <strong>and</strong> ethanol is produced. The expression of several sets of genes are repressed<br />
(Table 12.4). The respiratory pathways are inoperative even in the presence of oxygen.<br />
Gluconeogenesis is inhibited, as are pathways associated with the utilization of C 2<strong>and</strong> C 3<br />
compounds. Glucose repression is accompanied by changes in cell morphology. Thus, the<br />
biogenesisofmitochondria<strong>and</strong>peroxisomesisinhibited.Inaparallelseriesofevents,some<br />
ofthepre-formedenzymes,whichareencodedbythegenesofglucose-repressiblepathways,<br />
are inactivated by proteolysis. This phenomenon is termed glucose catabolite inactivation.<br />
The molecular basis of glucose catabolite repression is complex <strong>and</strong> not yet entirely<br />
elucidated. The evidence suggests that two parallel <strong>and</strong> antagonistic signalling pathways<br />
are involved (Stark, 1999). These pathways transmit the signal from the initial glucose<br />
trigger to the target genes <strong>and</strong> bring about the phenotypic changes associated with<br />
repression. In addition, in the absence of aglucose trigger, asignal is transmitted which<br />
brings about derepression. Thus, derepression is apositive function <strong>and</strong> is not brought<br />
about simply by the absence of arepressing signal.<br />
The initial receptor for the repressing glucose signal is unknown, however, one of the<br />
genes encoding for ahexokinase (HXK2) is involved. Thus, the repressing signal is not<br />
transmitted unless glucose is phosphorylated after uptake. No glycolytic genes after<br />
hexokinase are required for glucose repression <strong>and</strong> it does not occur in the absence of<br />
HXK2. There is circumstantial evidence that the Ras cyclic AMP-dependent protein<br />
kinase(cPKA)signalcascadeisalsoimplicated.Whenglucoseisaddedtoayeastculture<br />
growing oxidatively under conditions of glucose limitation, there is an immediate<br />
increaseinintracellularlevelsofcyclicAMP.ThesolefunctionofcyclicAMPinyeastis<br />
to activate cPKA. The activities of several enzymes, which take part in pathways<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 12.4 Genes, electron carriers <strong>and</strong> enzymes subject to glucose catabolite repression<br />
Gene Enzyme Metabolic Function<br />
CYC1 Cytochrome C Respiratory pathway<br />
COX6 Cytochrome oxidase<br />
QCR8 Ubiquinol cytochrome C oxidoreductase<br />
CIT1 Citrate synthase TCA cycle<br />
ACO1 Aconitase<br />
KGD1 -Ketoglutarate dehydrogenase<br />
MDH3 Malate dehydrogenase (cytosolic)<br />
MAL61 Maltose permease Disaccharide uptake <strong>and</strong> utilization<br />
MAL62 Maltase<br />
SUC2 Invertase<br />
GAL1 Galactokinase Galactose uptake <strong>and</strong> utilization<br />
GAL2 Galactose permease<br />
GUT1 Glycerol kinase Glycerol utilization<br />
GUT2 Glycerol 3-phosphate kinase<br />
FBP1 Fructose 1, 6-bisphosphatase Gluconeogenesis<br />
PCK1 Phosphoenolpyruvate carboxykinase<br />
ICL1 Isocitrate lysase Glyoxylate cycle<br />
MLS1 Malate synthase<br />
ADH2 Alcohol dehydroegenase II Ethanol utilization<br />
ACS1 Acetyl-CoA synthetase<br />
POX1 Acyl-CoA oxidase -oxidation<br />
POT1 3-Oxoacyl-CoA thiolase<br />
responsive to glucose repression, are influenced by cPKA. The effect of cPKA may be<br />
direct or indirect via other intervening kinases.<br />
The principal components of the signalling pathway mediated by Ras genes are shown<br />
in Fig. 12.14. The Ras proteins are subject to post-translational modification. This takes<br />
the form of addition of a GTP residue, a reaction catalysed by guanine-nucleotide<br />
exchange factor. The latter is the product of another gene (Cdc25p) whose expression is<br />
controlled by an, as yet, unidentified metabolite. The GTP-Ras protein is as an activator of<br />
adenylate cyclase <strong>and</strong> the activity of the latter catalyses the formation of cyclic AMP<br />
(3 1 ,5 1 -cyclic AMP).<br />
Cyclic-AMP-dependent protein kinase (PKA) comprises two regulatory sub-units<br />
(Bcy1P) <strong>and</strong> two catalytic sub-units (TpK). Cyclic AMP binds to the inactive PKA <strong>and</strong><br />
causes the regulatory sub-units to disassociate. Active PKA is liberated, which is then<br />
free to phosphorylate target enzymes <strong>and</strong> so transmit the glucose signal to the various<br />
target metabolic pathways <strong>and</strong> produce a response. Cyclic AMP is degraded by<br />
phosphatases. In addition, cPKA exerts feed back control on cAMP levels by<br />
phosphorylation of a component of the Ras transduction pathway.<br />
The antagonistic derepressing signal pathway is mediated by another kinase, Snf1p,<br />
levels of which increase rapidly in cells starved of glucose. Snf1p occurs as part of a<br />
multi-protein complex. In the presence of glucose the complex autophosphorylates <strong>and</strong><br />
lacks kinase activity. In the absence of glucose, a protein phosphatase (PP1) is activated<br />
by an unknown mechanism. This dephosphorylates the Snf1p-containing complex <strong>and</strong> as<br />
a result of a conformational shift the auto-inhibition is relieved <strong>and</strong> kinase activity is<br />
restored. Snf1p kinase apparently acts on another mediating protein, termed Mig1p. In the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glucose<br />
Cdc25p gene Ras gene Ira1 Ira2 genes<br />
Guanine-nucleotide<br />
exchange factor<br />
GDP-Ras<br />
protein<br />
GTP-RAS<br />
protein<br />
Cyr1p gene<br />
Adenylate cyclase<br />
ATP Cyclic AMP + PPi<br />
GAP protein<br />
GTP-ase<br />
PDE genes<br />
cAMP<br />
phosphodiesterase<br />
Inactive PKA Tpk.Bcy1P Bcy1P.cAMP<br />
Active PKA<br />
Tpk<br />
Phosphorylation of target genes<br />
Fig. 12.14 The Ras adenylate cyclase signal transduction system. The abbreviations are defined in<br />
the text.<br />
unphosphorylated form this protein binds to the promoter region of glucose-repressible<br />
genes <strong>and</strong> prevents their expression. When Mig1p is phosphorylated by Snf1p it<br />
dissociates from the target genes <strong>and</strong> repression is thereby lifted.<br />
Glucose catabolite repression is of importance in brewery fermentations. In the initial<br />
aerobic phase the presence of glucose ensures that metabolism is fermentative. In the<br />
later stages of fermentation, although the repressing signal may be absent, depending on<br />
the composition of the wort, anaerobiosis ensures that oxidative respiratory metabolism<br />
does not develop. The presence of glucose in wort prevents the utilization of the<br />
predominant sugar, maltose (Section 12.4.1). In all-malt worts this is of small importance<br />
since glucose concentrations are low, relative to maltose. Thus, during the early aerobic<br />
phase of fermentation glucose repression prevents the development of respiratory<br />
capacity <strong>and</strong> maltose utilization. However, this effect is transitory such that the<br />
disappearance of glucose <strong>and</strong> the onset of anaerobiosis are roughly coincident. At this<br />
stage anaerobiosis prevents the yeast from acquiring respiratory capacity <strong>and</strong> the<br />
disappearance of glucose removes the repressing effect such that maltose can be utilized.<br />
Caution should be exercised with worts containing glucose syrup adjuncts since the<br />
resultant prolonged repressing signal may prevent maltose utilization throughout much of<br />
the fermentation. In single-stage immobilized yeast reactors designed for primary<br />
fermentation glucose repression can have adverse effects. Thus, if throughput rates are<br />
too high the continuous addition of glucose in fresh wort can repress maltose utilization.<br />
This problem can be circumvented by the use of continuous systems containing two or<br />
more discrete stages. In such systems glucose is utilized in the first stage.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The glucose repression phenomenon is influenced by the availability of other nutrients.<br />
In the absence of acomplete growth medium, glucose-starved cells exposed to glucose<br />
exhibitatransientrepressionresponse.Followingadditionofthelimitingnutrient,asource<br />
of nitrogen for example, the typical glucose repression response is seen. This non-glucose<br />
response is identical to that mediated by cyclic AMP-dependent protein kinase (cPKA).<br />
However, the effect still proceeds in mutants lacking the regulatory sub-unit of cPKA.<br />
Ithas beensuggestedthatyeastsrequire additional controlsthatpreventthefull-blown<br />
glucose repression response in the absence of acomplete growth medium. When the<br />
latter condition is satisfied activation of protein kinase A(PKA) occurs <strong>and</strong> repression<br />
proceeds. The transmission of the initial trigger from the non-glucose nutrient to protein<br />
kinaseAoccursviaaregulatorysystemthatdoesnotinvolvecyclicAMP.Thissignalling<br />
mechanism is described as the fermentable-growth-medium induced pathway (Thevelein<br />
<strong>and</strong> Hohmann, 1995).<br />
ThePasteur effect isthe phenomenon whereby fermentation isinhibited by respiration<br />
or glycolytic rates decrease under aerobic conditions (Warburg, 1926). The energetic<br />
yield of respiration is more favourable than fermentation <strong>and</strong> furthermore, yields of<br />
cellular biomass per unit of sugar consumed are greater. It would be supposed, therefore,<br />
that under aerobic conditions yeast would preferentially use oxidative phosphorylation<br />
for the generation of energy <strong>and</strong> in consequence reduce rates of glycolysis. The<br />
magnitude of the Pasteur effect is dependent on the relative respiratory capacities of<br />
individual yeast strains. Both S. cerevisiae (a fermentative facultative anaerobe) <strong>and</strong><br />
C<strong>and</strong>ida tropicalis (a respiratory facultative anaerobe) growing exponentially on glucose<br />
under anaerobic conditions exhibit glycolytic rates of approximately 200 M.glucose<br />
consumed/min./g.dry wt. yeast. Under aerobic conditions, the glycolytic rate in S.<br />
cerevisiae is virtually unchanged, however, in C. tropicalis it decreases by more than<br />
90% to approximately 5 M.glucose consumed min 1 g 1 yeast dry weight (Gancedo <strong>and</strong><br />
Serrano,1989). In theformer yeast, acomparatively smallPasteur effect isobservedonly<br />
in glucose or nitrogen-starved stationary phase cultures.<br />
The mechanism of the Pasteur effect is obscure <strong>and</strong> possibly differs depending on<br />
cultural conditions <strong>and</strong> the nature of the yeast. In derepressed cells undergoing a<br />
transition from anaerobiosis to aerobiosis the effect may be due to simple competition for<br />
pyruvate. Since pyruvate dehydrogenase has ahigher affinity for pyruvate compared to<br />
pyruvate decarboxylase the presence of oxygen allows carbon to be diverted towards<br />
respiratorypathways<strong>and</strong>fermentationratesdecline.InyeastsuchasS.cerevisiaewithan<br />
inherentlylimitedrespiratorycapacity,thiseffectissmall.Decreaseinglycolyticratesby<br />
aerobiosis appears to involve feed-back control from oxidative phosphorylation. The<br />
mechanism is unknown although it has been suggested that glycolytic rates might be<br />
modulated by the effect of phosphate on the activity of phosphofructokinase (Gancedo<br />
<strong>and</strong> Serrano, 1989). The proposal is that under anaerobic conditions, flux through<br />
oxidative phosphorylation is reduced <strong>and</strong> this results in an increase in phosphate<br />
concentration. In turn, this activates phosphofructokinase <strong>and</strong> glycolytic activities are<br />
stimulated. Other mechanisms regulating sugar metabolism in various genera of nonbrewing<br />
yeasts have been discovered,for examplethe Custers<strong>and</strong>Kluyver effects.These<br />
are described elsewhere (Boulton <strong>and</strong> Quain, 2001).<br />
12.5.9 Ethanol toxicity <strong>and</strong> tolerance<br />
The biochemistry underlying the formation of ethanol from the catabolism of sugars is<br />
describedinSection12.5.5.Withregardtocommercialfermentations,inwhichethanolis<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
a major product of yeast metabolism, the rate of ethanol formation <strong>and</strong> the maximum<br />
concentration formed may be important considerations. In the case of the fermentation of<br />
high-gravity worts the ability of yeast to withst<strong>and</strong> high concentrations of ethanol is an<br />
influential factor in strain selection. <strong>Brewing</strong> yeast strains are exposed to ethanol<br />
concentrations typically in the range of 3±6% v/v. In high-gravity brewing ethanol<br />
concentration may be as high as 10% v/v. Wine yeasts are considered more ethanol<br />
tolerant than their brewing counterparts. In wine fermentations, final ethanol<br />
concentrations of 10±15% v/v are usual. In extreme cases such as Tokay wines <strong>and</strong> the<br />
rice `wine' sakeÂ, the fermentations generate in excess of 20% v/v ethanol.<br />
The ability to tolerate ethanol is usually considered to have a genetic basis. Exposure<br />
of yeast to very high ethanol concentrations results in the inhibition of growth <strong>and</strong><br />
ultimately death. There is no precise definition of ethanol tolerance in yeast. Some strains<br />
are more able than others to withst<strong>and</strong> the deleterious effects of ethanol. Of course, it is<br />
possible that intermediates of ethanol formation or other products of fermentation exert<br />
deleterious effects on yeast. In addition, in order to generate high ethanol concentrations<br />
during fermentation, it is necessary to provide a high initial concentration of fermentable<br />
sugar. In this case, the ability to grow under conditions of low water activity may be of<br />
greater or equal importance to ethanol tolerance, per se. In order to ferment concentrated<br />
worts it is essential that other nutrients are available in balanced quantities. This is an<br />
important consideration in high-gravity brewing where the injudicious use of sugar<br />
adjuncts may result in wort that is deficient in non-sugar nutrients.<br />
Ethanol inhibits yeast growth in a non-competitive manner. It does not have any<br />
specific inhibitory effect on glycolytic rate. Many reports describe morphological<br />
changes in response to ethanol exposure, for example, the development of cell surface<br />
invaginations <strong>and</strong> cell shrinkage (Pratt-Marshall et al., 2002). The toxic effects are<br />
apparently more severe when ethanol is generated endogenously compared to the same<br />
concentration added to the medium (Nagodawithana <strong>and</strong> Steinkraus, 1976). This<br />
observation prompted the suggestion that the effect was a consequence of intracellular<br />
accumulation of ethanol during fermentation. However, this assertion has been refuted. It<br />
appears that the plasma membrane is freely permeable to ethanol such that intracellular<br />
accumulation occurs only during very early fermentation (D'Amore et al., 1988). Other<br />
aliphatic alcohols also exert inhibitory effects. The severity of the inhibition is<br />
proportional to the chain-length of the molecule.<br />
Several mechanisms have been proposed by which the toxic effects of ethanol may be<br />
exerted. These include non-specific osmotic effects (Jones <strong>and</strong> Greenfield, 1987) <strong>and</strong> a<br />
number of specific target sites (D'Amore et al., 1990; Mishra, 1993). A high concentration of<br />
intracellular ethanol reportedly denatures some enzymes. Exposure to ethanol has a<br />
mutagenic effect on mitochondrial DNA as indicated by an increase in the occurrence of<br />
respiratory petites. The major site for ethanol toxicity appears to be the plasma membrane <strong>and</strong><br />
other intracellular membranes. Several effects have been observed which relate to membrane<br />
function. These include leakage of cellular components, abolition of membrane proton<br />
motive potential, inhibition of transport systems <strong>and</strong> alterations in membrane structure <strong>and</strong><br />
fluidity. The fact that the membrane is the primary target explains the observation that longer<br />
chain-length alcohols have enhanced toxicity. Thus, there is a concomitant increase in<br />
hydrophobicity <strong>and</strong> consequently easier interaction with membrane lipids.<br />
The observation that endogenously generated ethanol is more toxic than added ethanol<br />
at similar concentration suggests that other intermediates of ethanol biosynthesis might<br />
be influential. Potential c<strong>and</strong>idates include short chain fatty acids, higher alcohols,<br />
acetate <strong>and</strong> in particular acetaldehyde (Jones, 1987). The argument is perhaps most<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
persuasive in the case of acetaldehyde since reportedly, it is an order of magnitude more<br />
toxic to yeast than ethanol <strong>and</strong> is the immediate precursor of ethanol (Jones, 1989).<br />
However, the fact that acetaldehyde can accumulate in yeast cells, under some<br />
conditions, to agreater concentration than seen in fermentation <strong>and</strong> apparently with no<br />
toxic effect provides apowerful counter-argument (Stanley <strong>and</strong> Pament, 1993). It is<br />
perhaps most likely that intermediates of ethanol metabolism may contribute to ethanol<br />
toxicity in asynergistic fashion.<br />
Yeast cells exhibit various adaptations in response to exposure to ethanol. Ethanol<br />
elicits astress response such that there is aconcomitant acquisition of thermotolerance<br />
<strong>and</strong> barotolerance (Hisada et al., 2002). In addition, there is an increase in levels of<br />
intracellular trehalose (Mansure et al., 1994). Since trehalose is amembrane-stabilizing<br />
agent (Section 12.5.7), this response is explained. The lipid composition of membranes is<br />
altered such that there is an increase in the content of unsaturated fatty acids <strong>and</strong> 5,7unsaturated<br />
sterols <strong>and</strong> adecrease in saturated lipids. Of course, this response is not<br />
possible in abrewing fermentation where these syntheses are precluded by anaerobiosis.<br />
The toxic effects of ethanol are reportedly ameliorated by the presence of Mn-superoxide<br />
dismutase, the mitochondrial enzyme implicated in resistance to oxidative stress Costa et<br />
al., 1993). This enzyme is induced under oxidative conditions. Therefore, it would not be<br />
active under the conditions of abrewing fermentation (Section 12.6).<br />
The tolerance of yeast to ethanol can be influenced by manipulation of the growth<br />
medium. Predictably, supplementation of the medium with a source of unsaturated fatty<br />
acids is beneficial in this regard. A considerable body of evidence has now been amassed<br />
indicating that the addition to growth media of various metal ions provides protection<br />
against ethanol stress. In particular, calcium <strong>and</strong> magnesium have been reported to have<br />
beneficial effects (Dombek <strong>and</strong> Ingram, 1986; Ciessarova et al., 1996). It has been<br />
suggested that wort may not contain optimal concentrations of these metal ions (Walker<br />
et al., 1996; Rees <strong>and</strong> Stewart, 1997). This can be remedied by supplementation with<br />
magnesium to ensure that the ratio of this metal ion to calcium is always high. The<br />
mechanism by which metal ions exert protective effects is not clear.<br />
12.6 The role of oxygen<br />
Oxygen presents yeast with both an opportunity <strong>and</strong> a threat. Facultative anaerobic strains<br />
such as brewing yeasts have the ability to grow either oxidatively or fermentatively. The<br />
effects of glucose catabolite repression in Crabtree positive yeast are relieved in the<br />
absence of a glucose signal. However, development of complete respiratory competence<br />
requires the presence of oxygen. Thus, aerobiosis provides yeast with the opportunity of<br />
utilizing the energetically favourable oxidative route for energy production. However,<br />
oxidative metabolism is accompanied by the generation of potentially harmful reactive<br />
oxygen radicals. Consequently, yeast must have enzyme systems for removal of oxygen<br />
radicals <strong>and</strong> nullifying this potential threat.<br />
Three classes of genes in S. cerevisiae are recognized based on their response to oxygen<br />
tension (Zitomer <strong>and</strong> Lowry, 1992). Some genes are expressed only under anaerobic<br />
conditions. The role of many of these is unknown, however, some are involved in the<br />
assimilation of nutrients from the medium which otherwise require oxygen for their synthesis.<br />
Hypoxic genes are expressed strongly under micro-aerophilic conditions <strong>and</strong> are<br />
apparently required for the efficient utilization of low oxygen concentrations. The expression<br />
of more than 200 genes is required for aerobic respiratory growth (Tzagoloff <strong>and</strong> Dieckmann,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1990).Manyoftheseareexpressedonlyunderaerobicconditions.Theseincludecomponents<br />
of the electron transport chain such as ubiquinone <strong>and</strong> cytochrome oxidase.<br />
The signal pathway by which molecular oxygen exerts its effects upon metabolism is<br />
unknown. However, as described in the previous section in the hierarchy of signalling<br />
pathways, it plays asubordinate role to glucose repression. Haem is akey intermediate in<br />
the oxygen-sensing pathway. Formation of the haem precursor, protoporphyrin-IX is via<br />
the oxidation of coproporphyinogen-III. This is the rate-determining step in haem<br />
biosynthesis. Consequently, haem content <strong>and</strong> oxygen tension are directly related (De<br />
Winde <strong>and</strong> Grivell, 1993). In addition to its role as aprosthetic group in molecules such<br />
as cytochromes, haem is an effector metabolite in many pathways that utilise molecular<br />
oxygen. It is involved in the positive regulation of expression of genes encoding<br />
respiratory enzymes <strong>and</strong> those that play apart in protecting the cell against oxygen<br />
radicals. Conversely, haem represses the expression of several genes that are redundant<br />
under anaerobic conditions. These include some of those responsible for the synthesis of<br />
sterols <strong>and</strong> unsaturated fatty acids.<br />
<strong>Brewing</strong> yeasts do not develop respiratory competence under the conditions<br />
encountered in fermentation. Thus, in the aerobic phase of fermentation, respiratory<br />
pathways are repressed because of the presence of sugars. In late fermentation when the<br />
sugars have disappeared <strong>and</strong> their repressing effects are relieved, anaerobiosis prevents<br />
the induction of the respiratory enzymes.<br />
The majority of yeasts require oxygen for growth. In astudy of type species from 75<br />
genera, it was noted that only 23% could grow under anaerobic conditions on acomplex<br />
medium supplemented with ergosterol <strong>and</strong> asource of unsaturated fatty acids (Visser et<br />
al.,1990).Ofthese,S.cerevisiaewasexceptionalinthatitwascapableofrapidgrowthat<br />
low oxygen tension. Nevertheless, none of these yeasts, including S. cerevisiae, can grow<br />
under totally anaerobic conditions unless the medium is supplemented with asource of<br />
unsaturated fatty acids <strong>and</strong> sterols (Andreason <strong>and</strong> Stier, 1953a,b). These essential<br />
metabolites can be assimilated from the medium or synthesized de novo from<br />
carbohydrates. Synthesis requires the presence of molecular oxygen. Both of these are<br />
present in wort at the start of fermentation.<br />
In brewery fermentations, sterols <strong>and</strong> unsaturated fatty acids are synthesized during<br />
the aerobic phase. Cell proliferation during the anaerobic phase of fermentation dilutes<br />
the pre-formed pools of sterols <strong>and</strong> unsaturated fatty acids amongst daughter cells. On<br />
subsequent re-pitching, these lipids must be replenished hence the requirement for<br />
oxygenation of wort. Failure to provide sufficient oxygen is one of the prime causes of<br />
slow <strong>and</strong> sticking fermentations. The quantity of oxygen required for fermentation is<br />
strain-dependent. In an early study, ale strains were classified as requiring half air<br />
saturation, air saturation, oxygen saturation or more than oxygen saturation for satisfactoryfermentationperformance<br />
(Kirsop, 1974). Similar findingshave been reportedfor<br />
lager yeast strains (Jacobsen <strong>and</strong> Thorne, 1980). The explanation for these differences is<br />
related to the spectrum of sterols produced by individual yeast strains (Section 12.7.3).<br />
The fate of most of the oxygen utilized during the aerobic phase of fermentation is<br />
unknown. Theoretically 10% is utilized for sterol formation <strong>and</strong> 15% for the biosynthesis<br />
of unsaturated fatty acids (Kirsop, 1982). More than 50% is unaccounted for.<br />
During fermentative growth, yeast forms ATP via cytosolic substrate level<br />
phosphorylation since the mitochondrial oxidative electron transport is inoperative.<br />
Many essential energy-requiring enzyme systems are located within promitochondria, the<br />
undifferentiated organelles characteristic of fermentative yeast. Adenine nucleotides are<br />
transported between the cytosol <strong>and</strong> mitochondria via an ADP/ATP translocase. Three<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
isozymesofADP/ATP occurinS.cerevisiae,onethatisconstitutive,asecondinducedin<br />
respiratory cells <strong>and</strong> athird induced by anaerobiosis (Kolarov et al., 1990). It is proposed<br />
that the latter enzyme catalyses transfer of ATP from the cytosol into mitochondria<br />
during fermentative growth.<br />
Oxygen radicals are highly reactive species, which are implicated in damaging effects<br />
such as lipid peroxidation, mutagenesis <strong>and</strong> other degenerative changes associated with<br />
ageing <strong>and</strong> senescence. Yeast, in common with other cells, possesses protective<br />
mechanisms for removing oxygen radicals (Krems et al., 1995). The precursor of sterols,<br />
squalene, reportedly scavenges free radicals in mammalian cells (Kohno et al., 1995).<br />
Squalene accumulates in yeast under anaerobic conditions (Table 12.6 on page 447) <strong>and</strong><br />
it could fulfil a similar role. Similarly, reduced glutathione reacts with superoxide,<br />
hydrogen peroxide <strong>and</strong> larger hydroperoxides. The two major protective enzymes are<br />
superoxide dismutase (SOD) <strong>and</strong> catalase (Fridovich, 1986). These enzymes, acting in<br />
concert, convert the superoxide radical to oxygen <strong>and</strong> water.<br />
2 : O2 ‡ 2H‡ ! O2 ‡ H2O2<br />
2H2O2 ! 2H2O ‡ O2<br />
Yeast cells possess two superoxide dismutases, a cytosolic CuZn SOD <strong>and</strong> a<br />
mitochondrial Mn SOD. Mutants lacking both isozymes are hypersensitive to oxygen.<br />
The cytosolic CuZn SOD is constitutive <strong>and</strong> highly expressed in aerobic cultures <strong>and</strong><br />
those grown on non-fermentable substrates. The mitochondrial Mn SOD is induced by<br />
oxygen <strong>and</strong> it is not present in cells growing fermentatively. In brewing yeast, under nongrowth<br />
conditions during a transition from anaerobiosis to aerobiosis, there was a rapid<br />
increase in the specific activity of CuZn SOD. The specific activity of the Mn SOD<br />
increased only after several hours exposure to oxygen (Clarkson et al., 1991). The<br />
transition was accompanied by a decrease of 5±7% in the viability of the culture. It was<br />
concluded that the CuZn SOD was protective in anaerobic yeast, whereas the Mn SOD<br />
was important only in aerobic cultures.<br />
Two forms of catalase occur in yeast, a cytosolic form (catalase T, CTA1 gene) <strong>and</strong> a<br />
peroxisomal form (catalase A, CTT1 gene). Both are haemoproteins. CTA1 is induced by<br />
oxygen <strong>and</strong> growth on non-fermentable substrates such as fatty acids. It is repressed by<br />
glucose. CTT1 is also induced by oxygen <strong>and</strong> it requires haem for its expression. In<br />
addition, it is induced in response to stresses such as heat shock, low water activity <strong>and</strong><br />
oxidative stress (Dawes, 1999). These observations have resulted in the suggestion that<br />
catalase T is involved in hydrogen peroxide removal during the stationary phase,<br />
whereas, catalase A is protective towards sudden oxidative stress.<br />
12.7 Lipid metabolism<br />
Lipids are a diverse group of compounds that are characterized by the common property<br />
of sparing solubility in water but being soluble in organic solvents. They have important<br />
structural roles, especially in membranes. Frequently they form part of larger<br />
macromolecules where the hydrophobic nature of the lipid moiety confers specific<br />
properties. Many lipids have biological roles in signalling systems, as vitamins <strong>and</strong> in<br />
receptor sites on cell surfaces.<br />
S. cerevisiae, contains relatively modest levels of lipids, typically 5±15% of the cell<br />
dry weight (Rattray, 1988). The predominant classes of lipid are sterols (both free <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 12.5 Lipid composition (% total lipid) of various strains of S. cerevisiae (adapted from<br />
Rattray, 1988)<br />
Strain TAG DAG MAG FFA S SE PL Other<br />
1 7.4 ± ± 4.1 4.3 21.9 61.7 ±<br />
2 23.9 12.9 ± 1.9 ± 23.9 26.9 10.4<br />
3 8.2 1.3 0.9 1.9 6.6 53.3 27.7 ±<br />
4 7.4 0.3 0.8 0.8 4.7 45.4 38.3 ±<br />
5 9.6 5.8 ± 2.3 ± 25.3 56.9 ±<br />
6 40.4 4.3 ± 5.8 ± 19.6 29.9 ±<br />
7 12.6 4.9 ± ± 3.0 8.4 28.1 ±<br />
8 8.6 2.3 0.6 1.4 7.2 46.5 33.4 ±<br />
9 9.5 1.4 1.7 0.6 5.6 49.0 32.2 ±<br />
10 14.4 ± ± 1.2 1.7 25.4 52.0 5.4<br />
11 11.8 2.8 3.5 8.8 17.3 26.5 29.2 ±<br />
12 16.8 2.8 4.4 9.2 11.0 28.3 27.5 ±<br />
13 15.1 4.9 5.3 15.8 21.1 21.1 16.8 ±<br />
TAG, triacylglycerol; DAG, diacylglycerol; MAG, monoacylglycerol; FFA, free fatty acids; S, sterol; SE. steryl<br />
esters; PL, phosholipids.<br />
esterified), phospholipids <strong>and</strong> triacylglycerols. Fatty acids, either free or esterified as<br />
diacylglycerols <strong>and</strong>monoacylglycerolsmakeupmostoftheremaininglipid(Table12.5).<br />
The predominant saturated fatty acids are palmitic (16:0) <strong>and</strong> stearic (18:0) with smaller<br />
amounts of myristic (14:0) <strong>and</strong> lauric (12:0). Unsaturated fatty acids are mainly<br />
palmitoleic (16:1), oleic (18:1) <strong>and</strong> linoleic (18:2).<br />
12.7.1 Fatty acid metabolism<br />
Fatty acids are synthesized from acetyl-CoA. The latter may arise via several routes.<br />
Principally, it is derived from glucose catabolism from pyruvate, directly or via<br />
acetaldehyde, acetate <strong>and</strong> acetyl-CoA synthetase. It is formed from the catabolism of<br />
amino acids, leucine, lysine, tryptophan, tyrosine <strong>and</strong> phenylalanine. In addition, acetyl-<br />
CoA is the end-product of the -oxidation pathway for the degradation of fatty acids.<br />
The biosynthetic pathway to fatty acids involves the action of two enzyme systems,<br />
acetyl-CoA carboxylase <strong>and</strong> the fatty acid synthase complex. Acetyl-CoA carboxylase<br />
catalysestheconversionofacetyl-CoAintomalonyl-CoA.Thereactioniscomplex<strong>and</strong>is<br />
drivenbythebreakdownofATP.Theadditionalcarbonatomisderivedfrombicarbonate<br />
ion in areaction involving the coenzyme biotin.<br />
Enzyme-biotin‡ATP‡HCO3 enzyme-biotin-CO2 ‡ADP‡Pi<br />
Enzyme-biotin-CO2 ‡acetyl-CoA malonyl-CoA‡enzyme<br />
Acetyl-CoA carboxylase is encoded by the gene ACC1. Asecond gene ACC2 encodes a<br />
biotin-apoprotein ligase which catalyses the addition of the biotin moiety to the ACC1<br />
gene product <strong>and</strong> converts it from the inactive apo- to the active holo- form. Acetyl-CoA<br />
carboxylase is considered the rate-determining step in fatty acid biosynthesis. It is<br />
activated allosterically by citrate <strong>and</strong> isocitrate.<br />
Fatty acid synthase (FAS) is amultienzyme complex that catalyses asequence of<br />
reactions in each cycle of which afatty acid is lengthened by two carbon atoms. (Fig.<br />
12.15). The two carbon atoms are donated by malonyl-CoA deriving from the activity of<br />
acetyl-CoA carboxylase. The FAS complex contains acyl carrier protein (ACP) which<br />
contains 1 0 -phosphopantetheine as a prosthetic group. During the sequence of reactions,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
HOOC CH2 CO S CoA<br />
HCO3 –<br />
CH3 CO S CoA<br />
Acetyl-CoA<br />
ATP<br />
NADP +<br />
NADPH + H +<br />
CO2 ACP<br />
Acetoacetyl-S-ACP<br />
H2O<br />
NADPH + H +<br />
Crotonyl-S-ACP<br />
NADP +<br />
CoA<br />
Malonyl-CoA<br />
Acetyl-S-ACP Malonyl-S-ACP<br />
D-β-Hydroxybutyryl-S-ACP<br />
Butyryl-S-ACP<br />
FAS 7<br />
cycles<br />
CoASH<br />
intermediates <strong>and</strong> substrates are bound to the SH-group of ACP via thioester linkages.<br />
Activation of FAS requires activity of phosphopantetheinyl transferase, which attaches<br />
the 1 0 -phosphopantetheine arm to a serine residue in ACP.<br />
Each cycle of reactions involves a priming step in which acyl groups are transferred<br />
from CoA to ACP. This is followed by a condensation reaction where the acyl <strong>and</strong><br />
malonyl groups are combined with the loss of the third carbon atom as CO 2. Finally, there<br />
are two reductive steps involving the coenzyme, NADPH+H + <strong>and</strong> an intermediate<br />
dehydration step. Seven successive cycles of the FAS complex leads to the formation of<br />
the C16:0 fatty acid ester, palmitoyl-CoA.<br />
Acetyl-CoA ‡ 7malonyl-CoA ‡ 14 NADPH ‡ 14H ‡<br />
CH 3<br />
CH 3<br />
CO CH2 CO S ACP<br />
CH3 CHOH CH2 CO S ACP<br />
CH CH CO S CoA<br />
CH3 CH2 CH2 CO S CoA<br />
Acetyl CoA carboxylase<br />
! palmitoyl-CoA ‡ 7 CoASH ‡ 7CO2 ‡ 7H2O ‡ 14 NADP ‡<br />
Unsaturated fatty acids (UFA) are produced by the action of the enzyme, -9- fatty<br />
acid desaturase, which inserts a double bond into fatty acid under aerobic conditions.<br />
Multiple desaturation reactions result in the synthesis of di-<strong>and</strong> trienoic acids such as<br />
linoleic (18:2) <strong>and</strong> linolenic (18:3) acids. The desaturase enzyme is encoded by the gene<br />
1<br />
–<br />
2O2<br />
Palmitoyl-S-ACP Palmitoyl-CoA Palmitoleic-CoA<br />
H2O<br />
Fatty acid desaturase<br />
Fatty acid<br />
synthase<br />
(1 cycle)<br />
Fig. 12.15 Pathway for the biosynthesis of saturated <strong>and</strong> unsaturated fatty acids (ACP ˆ acyl<br />
carrier protein).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
OLE1. The enzyme is located in the endoplasmic reticulum <strong>and</strong> its activity is linked to an<br />
NADPH-dependent cytochrome b5 reductase. Activity of the desaturase is regulated by<br />
the presence of exogenous UFA. Unsaturated fatty acids apparently cause rapid<br />
degradation of the OLE1 mRNA (Gonzales <strong>and</strong> Martin, 1996).<br />
Fatty acids, both saturated <strong>and</strong> unsaturated, eventually become located in membranes<br />
where they have structural roles. The relative chain lengths <strong>and</strong> degrees of unsaturation<br />
are influenced by environmental conditions. For example, as the growth temperature is<br />
reduced there is a need to maintain membrane fluidity. This is achieved by increasing the<br />
degree of unsaturation of the fatty acids in membrane lipid components. In addition to<br />
desaturases, S cerevisiae also possesses a membrane-bound fatty acid elongation system,<br />
encoded by the ELO genes. These are independent of the FAS system <strong>and</strong> are used to<br />
elongate short chain fatty acids (C10±C12) assimilated from the medium to C16±C18.<br />
The elongation reactions occur prior to insertion of fatty acids into membranes<br />
(Schweizer, 1999).<br />
It would be predicted that traditional lager fermentations performed at relatively low<br />
temperatures would have an increased requirement for UFA as a consequence of the<br />
correlation of the latter <strong>and</strong> membrane fluidity. This possibility appears not to have been<br />
explored in detail although it has been reported that UFAs are essential for satisfactory<br />
fermentation performance. This effect was ascribed to the requirement for UFA for<br />
proper mitochondrial development <strong>and</strong> function (O'Connor-Cox et al., 1993). UFAs are<br />
reported to influence the formation of esters (12.10.4). Linoleic acid reduces the<br />
formation of acetyl esters, reportedly by the repression of ATF1, the gene encoding<br />
alcohol acetyltransferase (Fuji et al., 1997).<br />
Fatty acid biosynthesis occurs in the cytosol, although there is evidence that yeast may<br />
possess a second mitochondrial fatty acid synthase (Schneider et al., 1997). Fatty acid<br />
degradation in S. cerevisiae occurs in peroxisomes via -oxidation. The process consists<br />
of a cyclic two-carbon shortening of fatty acids catalysed by a series of enzymes encoded<br />
by FOX genes. The first <strong>and</strong> rate limiting step in the pathway is catalysed by an acyl CoA<br />
oxidase <strong>and</strong> converts fatty acyl-CoA esters to trans 2, 3-dehydroacyl-CoA esters <strong>and</strong><br />
hydrogen peroxide. The reaction utilizes flavin adenine nucleotide (FAD) as cofactor <strong>and</strong><br />
this transfers electrons directly to molecular oxygen. Hydrogen peroxide is degraded by<br />
the peroxisomal catalase A. In the next sequence of reactions, catalysed by a<br />
multifunctional enzyme, trans 2, 3-dehydroacyl-CoA esters are successively modified<br />
by the action of trans 2-enoyl CoA hydratase <strong>and</strong> 3-hydroxyacyl CoA dehydrogenase.<br />
Finally, 3-ketoacyl CoA thiolase releases a molecule of acetyl-CoA leaving a residual<br />
acyl-CoA chain two carbon atoms shorter than the original.<br />
The three enzymes of -oxidation are induced during growth on fatty acids under<br />
aerobic conditions. In addition, the first enzyme, acyl-CoA oxidase is repressed by glucose.<br />
Fatty acids are transported from the cytosol into peroxisomes via transporters specific for<br />
medium chain length <strong>and</strong> long chain length fatty acids. The acetyl-CoA generated by -<br />
oxidation is transported back into the cytosol via a carnitine/acetylcarnitine shuttle system<br />
or as citrate. In the latter case, the acetyl-CoA is acted upon by a peroxisomal citrate<br />
synthase. In view of the susceptibility to glucose repression <strong>and</strong> requirement for oxygen, -<br />
oxidation is unlikely to play any part in brewery fermentations.<br />
12.7.2 Phospholipids<br />
Fatty acyl-CoA esters are trans-esterified to form acylglycerols via the phosphatidic acid<br />
pathway. The free 1, 2-hydroxyl groups of glycerol 3-phosphate are acylated by two<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
molecules of fatty acyl-CoA to yield a phosphatidic acid. The phosphate is removed from<br />
the latter by a phosphatase. The resultant diacylglycerol reacts with a third molecule of<br />
fatty acyl-CoA ester to form a triacylglycerol. Two phosphatidic acid phosphatases occur,<br />
one microsomal <strong>and</strong> another that is located in mitochondria. The microsomal enzyme is<br />
subject to regulation by the Ras cyclic AMP-dependent protein kinase. In non-oleaginous<br />
yeasts, such as brewing strains, triacylglycerols are synthesized only when the fatty acid<br />
requirement for phospholipids is satisfied. Triacylglycerols accumulate with steryl esters<br />
in cytosolic lipid granules. The fatty acids may be mobilized by the action of lipases,<br />
providing an alternative source of fatty acids for phospholipids synthesis.<br />
The major phospholipids in yeast are glycerophospholipids. In S. cerevisiae the<br />
predominant types are phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol<br />
<strong>and</strong> phosphatidylserine (Rattray, 1988). Phospholipids consist of a molecule of<br />
glycerol in which one of the hydroxyl groups is esterified to phosphoric acid <strong>and</strong> the other<br />
two to fatty acids. The phosphate head group form ester linkages with molecules such as<br />
amino acids <strong>and</strong> alcohols to give the phospholipids detailed above. The combination of<br />
polar head group <strong>and</strong> hydrophobic fatty acyl chains confers amphipathic properties on<br />
phospholipids, which makes them suitable for insertion into membranes. Inositol<br />
phospholipids have roles in cellular signalling systems <strong>and</strong> they are involved in the<br />
regulation of protein sorting across intracellular membranes.<br />
Like acylglycerol lipids, phospholipids are synthesized from phosphatidic acids (Fig.<br />
12.16). Phosphatidylcholine <strong>and</strong> phosphtidylethanolamine are formed from diacylglycer-<br />
Acyl CoA<br />
CoASH<br />
Choline<br />
Pi<br />
CTP<br />
Pi<br />
CDP-choline Lysophosphatidic acid<br />
Phosphatidyl Acyl CoA<br />
choline<br />
CoASH<br />
CMP Pi H2O Phosphatidic<br />
acid<br />
CTP Pi<br />
Diacylglycerol<br />
Triacylglycerol<br />
CMP<br />
CDP-ethanolamine<br />
Phosphoethanolamine<br />
ADP<br />
ATP<br />
CTP<br />
Ethanolamine<br />
Glycerol 3-phosphate<br />
Phosphatidyl<br />
ethanolamine<br />
Acyl CoA<br />
CoASH<br />
Pi<br />
CDP-diacylglycerol<br />
Phosphatidylinositol<br />
Phosphatidyl<br />
glycerol<br />
L-serine<br />
CMP<br />
Phosphatidylserine<br />
Myo-inositol<br />
CMP<br />
Fig. 12.16 Synthesis of phospholipids <strong>and</strong> acyl-glycerol lipids via the phosphatidic acid pathway.<br />
CTP ˆ cytidine 5 0 -triphosphate, CDP ˆ cytidine 5 0 -diphosphate,<br />
CMP ˆ cytidine 5 0 -monophosphate.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ols. Phosphatidylglycerol (cardiolipin), phosphatidylserine <strong>and</strong> phosphatidylinositol are<br />
synthesized from the common precursor, CDP-diacylglycerol. Phospholipids are<br />
synthesized in reactions that involve the carrier molecule, cytidine 5 0 -diphosphate<br />
(CDP). Phospholipids are degraded by specific phospholipases that remove the fatty acyl<br />
groups from the glycerol molecule.<br />
12.7.3 Sterols<br />
Sterols, like unsaturated fatty acids, are formed during the initial phase of wort<br />
fermentationwhenoxygenisavailable.Themostabundantsterolinyeastisergosterol.In<br />
brewing strains, smaller quantities of zymosterol, episterol, lanosterol, <strong>and</strong> fecosterol can<br />
also be detected (Table 12.6). Sterols are essential components of cell membranes where,<br />
in conjunction with phospholipids, they confer fluidity. Sterols are most abundant in the<br />
plasma membrane, where they occur in the free form. Some 90% of the sterol in the<br />
plasma membrane is ergosterol. The smaller quantities of other sterols are probably pools<br />
of intermediates in the ergosterol biosynthetic pathway. Membranes surrounding<br />
intracellular organelles contain smaller quantities of sterols. Lipid granules, which<br />
contain cellular reserves of triacylglycerol, also contain sterols esterified to fatty acids.<br />
The starting point for sterol synthesis is acetyl-CoA. The latter may arise from the<br />
catabolism of wort sugars via glycolysis. However, in the early phase of brewery<br />
fermentations membrane function of pitching yeast may be impaired by sterol depletion.<br />
Consequently, it has been suggested that the carbon <strong>and</strong> energy for sterol synthesis could<br />
be supplied by mobilization of glycogen reserves. Alinear correlation between glycogen<br />
breakdown <strong>and</strong> sterol synthesis during the aerobic phase of wort fermentation has been<br />
demonstrated (Quain <strong>and</strong> Tubb, 1982). Yeast can utilize exogenous sterols. This occurs<br />
only under anaerobic conditions when de novo synthesis is precluded by the absence of<br />
oxygen. This phenomenon has been termed aerobic exclusion (Parks <strong>and</strong> Casey, 1995).<br />
By inference, any sterols present in wort would not be utilized during the initial aerobic<br />
phase of fermentation.<br />
In the first part of the sterol biosynthetic pathway, three acetyl units are combined to<br />
form a molecule of mevalonate (Fig. 12.17). Mevalonate is then converted to 3isopentenyl<br />
pyrophosphate in a sequence of phosphorylation reactions in which three<br />
molecules of ATP are hydrolysed. Thus, sterol synthesis is expensive in terms of the<br />
expenditure of metabolic energy. A molecule of 3-isopentenyl pyrophosphate <strong>and</strong> its<br />
isomer, 3-3-dimethylallyl pyrophosphate then condense with a loss of pyrophosphate to<br />
form the monoterpene derivative, geranyl pyrophosphate. This reacts with another<br />
molecule of 3-isopentenyl pyrophosphate to form the sesquiterpene derivative, farnesyl<br />
Table 12.6<br />
data)<br />
Squalene <strong>and</strong> sterol composition of lager pitching yeast (S. C. P. Durnin, unpublished<br />
Component % Dry wt.<br />
Squalene 1.2<br />
Ergosterol 0.095<br />
Lanosterol 0.055<br />
4, 4-Dimethylzymosterol 0.022<br />
Zymosterol 0.01<br />
Ergosta-7, 22-dienol 0.009<br />
Dihydroergosterol 0.007<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Acetyl-CoA<br />
PPi<br />
Geranyl pyrophosphate<br />
PPi<br />
Ubiquinone Farnesyl pyrophosphate<br />
PPi<br />
Acetyl-CoA<br />
Acetoacetyl-CoA<br />
CoA<br />
Hydroxymethylglutaryl-CoA<br />
NADPH<br />
CoA<br />
Mevalonic acid<br />
3ATP<br />
3ADP<br />
3-Isopentenyl pyrophosphate<br />
Squalene<br />
pyrophosphate. This metabolite is also a precursor for haem <strong>and</strong> ubiquinone biosynthesis.<br />
Two molecules of farnesyl pyrophosphate condense to give presqualene pyrophosphate,<br />
which is then reduced by NADPH with a loss of pyrophosphate to form squalene.<br />
The initial part of the sterol biosynthetic pathway is an anaerobic process. The<br />
biosynthesis of ergosterol from squalene requires molecular oxygen. The precise steps of<br />
the biosynthetic pathway differ in individual strains but a generalized scheme is shown in<br />
Fig. 12.17. Oxygen is utilized in the first step in which squalene is converted to the<br />
epoxide, 2, 3-oxidosqualene. Some of the subsequent steps utilize cytochrome P450<br />
oxygenases, which also require molecular oxygen.<br />
Sterol biosynthesis is highly regulated. The key step in the early part of the pathway is<br />
the formation of mevalonate from hydroxymethylglutaryl-CoA (HMG-CoA), catalysed<br />
by HMG-CoA reductase. This step is thought to be the rate-limiting step in the<br />
biosynthesis of sterols <strong>and</strong> other isoprenoids (Schweizer, 1999). Two isozymes of HMG-<br />
CoA reductase occur in yeast, encoded by the genes HMG1 <strong>and</strong> HMG2. Expression of<br />
both genes is regulated by oxygen. HMG1 is expressed to the greatest extent under highly<br />
aerobic conditions, whereas, HMG2 is strongly expressed under conditions of hypoxia.<br />
The transcription of both enzymes is activated by haem. The modest sterol concentrations<br />
synthesized by brewing yeast during fermentation probably result from the activity of<br />
O2<br />
Lanosterol Squalene epoxide<br />
Ignosterol<br />
4,4-Dimethylzymosterol<br />
CoA<br />
NADPH<br />
3,3-Dimethylallyl<br />
pyrophosphate<br />
Zymosterol Fecosterol Episterol Ergosterol<br />
Fig. 12.17 Biosynthetic pathway of ergosterol from acetyl-CoA.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
HO
HMG2 alone. The relatively high sterol concentrations formed in brewing strains of S.<br />
cerevisiae growing under derepressed conditions probably require the activity of both<br />
HMG1 <strong>and</strong> HMG2.<br />
Regulation of carbon flow up to squalene requires apartitioning of flux between the<br />
formation of sterols <strong>and</strong> the biosynthesis of other isoprenoids. This is accomplished by<br />
controloftheactivitiesoftheenzymesinthe earlypartofthe sterolbiosyntheticpathway<br />
<strong>and</strong> the expression of the genes encoding them. Squalene synthase is the first committed<br />
step in sterol synthesis. It appears that ergosterol <strong>and</strong> other late intermediates in the sterol<br />
pathway regulate the biosynthesis of ergosterol by feedback control mechanisms<br />
controlling the activities of enzymes earlier in the pathway.<br />
Sterols are reportedly synthesized in the microsomal fraction of the endoplasmic<br />
reticulum (Schweizer, 1999). Other have asserted that a separate partitioned sterol<br />
biosynthetic pathway may occur in mitochondria (Casey et al., 1992). To complicate<br />
matters further, sterol esters are located in cytosolic lipid granules. These membrane<br />
bound structures also contain triacylglycerols. In addition, sterol C-24 methyltransferase,<br />
a late enzyme of the sterol biosynthetic pathway is also found here as well as<br />
intermediates of sterol biosynthesis (Schweizer, 1999). The implication is that there may<br />
be more than one site of synthesis <strong>and</strong> in any case, an intracellular sterol transport system<br />
must exist. It is likely that under conditions of aerobic exponential growth, sterols are<br />
synthesized in the endoplasmic reticulum, transported to appropriate sites <strong>and</strong><br />
incorporated into proliferating membranes. In stationary phase cells, where membrane<br />
proliferation has ceased, sterols are esterified by acyl-CoA: sterol acyltransferases <strong>and</strong><br />
stored in lipid granules. If the conditions change such that growth recommences, steryl<br />
esters are transported to the sites of membrane growth <strong>and</strong> there they are cleaved by<br />
steryl ester hydrolases. The resultant free sterol may then be used for membrane<br />
synthesis.<br />
Inbrewingfermentationstheextentofsterolsynthesisinyeastismodestcomparedto<br />
derepressed cells. Pitching yeast typically contains 0.1±0.2% of the cell dry weight as<br />
sterol. This increases to approximately 1% of the cell dry weight at the end of the<br />
aerobic phase of fermentation. The same yeast grown aerobically under derepressing<br />
conditions contains approximately 5% of the cell dry weight as sterol (Quain <strong>and</strong> Tubb,<br />
1982).<br />
12.8 Nitrogen metabolism<br />
In wort, the major sources of nitrogenous nutrients are amino acids <strong>and</strong> ammonium ions.<br />
It would be supposed that amino acids might be incorporated directly into proteins <strong>and</strong><br />
other macromolecules. However, in brewing strains of S. cerevisae, growing<br />
fermentatively on wort, amino acids are catabolized (Jones <strong>and</strong> Pierce, 1970). By<br />
inference, amino acids that are required for the biosynthesis of other macromolecules<br />
must themselves be synthesized. Amino acids are usually degraded by long <strong>and</strong><br />
convoluted pathways. These pathways provide essential precursors for other nitrogencontaining<br />
cellular constituents such as purines <strong>and</strong> pyrimidines. This anabolic<br />
requirement presumably explains the preference for degradation. The oxidative<br />
catabolism of amino acids ultimately removes the amino group. Providing the yeast is<br />
growing under derepressed <strong>and</strong> aerobic conditions the resultant carbon skeletons can be<br />
fed into oxidative energy-producing pathways or provide precursors for gluconeogenesis<br />
(Fig. 12.18). In this way nitrogen <strong>and</strong> sugar metabolism are coupled <strong>and</strong> coordinated.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH3 CO CH2 CO S CoA<br />
Acetoacetyl-CoA<br />
Phenylalanine<br />
(4.43)<br />
Tyrosine (4.48)<br />
Lysine (4.40)<br />
Alanine (4.24)<br />
Threonine (4.46)<br />
Glycine (4.35)<br />
Serine (4.45)<br />
Cysteine (4.31)<br />
Pyruvate<br />
Acetyl-CoA<br />
Leucine (4.39)<br />
Trytophan (4.47)<br />
Isoleucine (4.38)<br />
Citrate<br />
Arginine (4.28)<br />
Histidine (4.36)<br />
Glutamine (4.34)<br />
Proline (4.44)<br />
Glutamate (4.33)<br />
2-Oxoglutarate<br />
Succinyl-CoA<br />
Fumarate<br />
Oxaloacetate<br />
Aspartate (4.29)<br />
Asparagine (4.30)<br />
Alternatively, under fermentative conditions some of the products of amino acid<br />
catabolism may be released into the medium. As aresult of the fermentative growth of<br />
yeast on wort, these may contribute to beer flavour. The carbon skeletons for amino acid<br />
synthesis may be derived from glycolysis. They may also be formed from the products of<br />
the degradation of other amino acids or ammonia may supply nitrogen for the amino<br />
groups. Thus, S. cerevisiae can utilize ammonia <strong>and</strong> most individual amino acids as sole<br />
sources of nitrogen. The biosynthetic pathways are not usually asimple reversal of those<br />
used for amino acid degradation. Key reactions in amino acid metabolism are<br />
transaminations where the -amino group of an amino acid is transferred to the -<br />
carbon atom of an -keto acid. The latter is usually -ketoglutaratic acid. For example,<br />
the reaction shown below in which the amino group of L-aspartate is transferred to -<br />
ketoglutarate to form L-glutamate.<br />
L-aspartate‡ -ketoglutarate oxaloacetate‡L-glutamate<br />
Isoleucine<br />
(4.38)<br />
Methionine<br />
(4.41)<br />
Valine<br />
(4.49)<br />
Phenylalanine<br />
(4.43)<br />
Tyrosine<br />
(4.48)<br />
Fig. 12.18 Fate of carbon skeletons of amino acids following oxidative degradation. In some<br />
cases, as indicated, some amino acids contribute to the pools of more than one intermediate of the<br />
main sugar oxidative pathways. The numbers in parentheses indicate where the molecular structures<br />
may be found.<br />
These reactions are reversible <strong>and</strong> many amino acids participate inthem. The deaminated<br />
amino acid is converted to the corresponding -keto acid analogue. These -keto acids<br />
are precursors of other metabolic by-products, which contribute to beer flavour, for<br />
example higher alcohols <strong>and</strong> esters (Section 12.10). In addition, they are themselves used<br />
to synthesize the corresponding amino acids, as required. The -keto acid analogues of<br />
some amino acids are shown in Table 12.7.<br />
Wort amino acids have been classified based on the relative contribution of their<br />
corresponding -keto acid analogues to the development of abalanced spectrum of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 12.7 Amino acids <strong>and</strong> their corresponding -keto acid analogues<br />
Amino acid -keto acid analogue<br />
Alanine Pyruvic acid<br />
Aspartate Oxaloacetic acid<br />
Glutamate -ketoglutaric acid<br />
Isoleucine -keto- -methylvaleric acid<br />
Leucine -ketoisocaproic acid<br />
Phenylalanine Phenylpyruvic acid<br />
Serine Hydroxypyruvic acid<br />
Tyrosine Hydroxyphenylpyruvic acid<br />
Valine -ketoisovaleric acid<br />
Table 12.8 Classification of amino acids based on the essential nature of the corresponding -<br />
keto acid analogue (from Jones <strong>and</strong> Pierce, 1970)<br />
Class 1 Class 2 Class 3<br />
Aspartate Isoleucine Lysine<br />
Asparagine Valine Histidine<br />
Glutamate Phenylalanine Arginine<br />
Glutamine Glycine Leucine<br />
Threonine Alanine<br />
Serine Tyrosine<br />
Methionine<br />
Proline<br />
flavour compounds (Jones <strong>and</strong> Pierce, 1970). The initial concentrations in wort of<br />
members of Class 1 (Table 12.8) were considered relatively unimportant since they<br />
could be either assimilated from wort or synthesised de novo. Thus, there is no<br />
shortage of intermediates <strong>and</strong> the provision of precursors for the synthesis of other<br />
essential products of metabolism. The initial concentration of Class 2 amino acids was<br />
considered crucial since in the later stages of fermentation synthesis of these from<br />
sugars was repressed. Consequently, the -keto acid analogues derived from amino<br />
acid degradation became the major sources of these amino acid carbon skeletons. A<br />
shortage of the latter in late fermentation would be predicted to have large effects on<br />
the metabolism of related by-products <strong>and</strong> by inference beer quality. Class 3 amino<br />
acids were considered to be derived exclusively from wort. Deficiencies in these would<br />
be expected to restrict the synthesis of compounds derived from their -keto acid<br />
analogues.<br />
Glutamate <strong>and</strong> the ammonium ion are central to nitrogen metabolism since they link both<br />
catabolism <strong>and</strong> anabolism. The key enzymes are glutamine synthetase, glutamate synthase<br />
(GOGAT) <strong>and</strong> glutamate dehydrogenase. These catalyse the reactions shown following.<br />
NH4 ‡ ‡ glutamate ‡ ATP glutamine ‡ ADP ‡ Pi<br />
Glutamine synthetase<br />
glutamine ‡ -ketoglutarate ‡ NADPH ‡ H ‡ ! 2 glutamate ‡ NADP ‡<br />
Glutamate synthetase<br />
glutamate ‡ NAD…P† ‡ H2O -ketoglutarate ‡ NH4 ‡ ‡ NAD…P†H<br />
Glutamate dehydrogenase<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glutamate <strong>and</strong> glutamine are the major sources of cellular nitrogen. Ammonium ions are<br />
converted to glutamate by glutamate dehydrogenase. Some of the glutamate is then<br />
converted toglutaminebyglutaminesynthetase. Thelatter enzyme isrequiredfor growth<br />
on any nitrogen source other than glutamine. Glutamine is an essential precursor for the<br />
biosynthesis of other amino acids as well as purines <strong>and</strong> pyrimidines. Yeasts have three<br />
glutamate dehydrogenases, one which is NAD + -linked (Gdh2). The other two require<br />
NADP + ascofactor, termed Gdh1 <strong>and</strong> Gdh3, respectively. The Gdh1 gene product has a<br />
catabolic function in the formation of glutamate, as described already. The Gdh2 enzyme<br />
iscatabolicinnature<strong>and</strong>servestoproduceammonia<strong>and</strong> -ketoglutarateincellsutilising<br />
glutamate. The role of the Gdh3 enzyme is not clear, although it has been speculated that<br />
it is involved in anitrogen-sensing pathway (Wilkinson et al., 1996).<br />
Glutamate synthase performs the same function as the Gdh1, NADP + -dependent<br />
glutamate dehydrogenase in that it produces glutamate in conjunction with glutamine<br />
synthetase. Based on differences in kinetic properties it has been claimed that the<br />
GOGAT system ishigh affinity <strong>and</strong> low activity, whereasthe Gdh1 system isthe reverse.<br />
The enzyme systems described are regulated by the availability of nitrogen. Thus, Gdh1<br />
<strong>and</strong> Gdh2 activities are inversely related depending on the availability of ammonia.<br />
Glutamine synthetase activity is induced when glutamate is the sole source of nitrogen<br />
but low in the presence of glutamine.<br />
The regulation of amino acid metabolism during growth of S. cerevisiae on an<br />
undefined medium such as wort is complex. Intermediates <strong>and</strong> end-products of the<br />
pathways leading to <strong>and</strong> from individual amino acids exert control by feed-back<br />
mechanisms. Inaddition, there are global regulatory mechanisms.Starvation ofanyone of<br />
several amino acids induces many of the enzymes required for the biosynthesis of several<br />
amino acids. The phenomenon involves the increased transcription of more than 40 genes<br />
<strong>and</strong>it hasbeen termed general amino acid control (Hinnebusch, 1997). With respect to the<br />
utilization of nitrogen sources, the presence of ammonia or glutamine causes the<br />
repressionoftheenzymes requiredforthecatabolism ofotheraminoacids.Thisprocessis<br />
termed nitrogen catabolite repression (Wiame et al., 1985). Other metabolic controls may<br />
be overlaid on those solely driven by the nitrogen source. For example, in derepressed<br />
yeast, biosynthesis of serine occurs via glycine, derived from glyoxylate. However, in<br />
repressed yeast as is the case in fermentation, the glyoxylate cycle is not functional<br />
(12.5.6) <strong>and</strong> serine may be formed from glycolysis via 3-phosphoglycerate.<br />
Glutamine, with other amino acids, provides nitrogen groups for the synthesis of<br />
purines, pyrimidines <strong>and</strong> N-acetylglucosamine. The latter is used in the synthesis of<br />
chitin, astructural component of yeast cell walls. Amino acids are the building blocks of<br />
proteins. Areview of protein synthesis in yeast may be found in Tuite (1991). Yeast cells<br />
possess anumber of enzymes responsible for the hydrolytic degradation of proteins,<br />
termed proteinases. These enzymes fulfil intracellular roles since S. cerevisiae does not<br />
utilize exogenous proteins. Some proteinases have abroad specificity <strong>and</strong> are involved in<br />
long-term protein turnover. Others have very specific substrates <strong>and</strong> have regulatory<br />
functions. The latter group catalyse reactions in which target proteins are modified by<br />
partial proteolysis such that they are reversibly activated or inactivated. These processes<br />
are distinct from the phenomenon of catabolite inactivation (Section 12.5.6).<br />
The major site for proteolysis is the cell vacuole. Much of the regulation of both<br />
specific <strong>and</strong> non-specific proteolysis involves the sequestration of target proteins into<br />
vacuoles where they are exposed to proteinases. The majority of proteins are turned over<br />
very slowly. A small proportion is subject to rapid turnover. The latter group includes<br />
proteins required only under specific conditions, for example, cyclins.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The tasks of elimination of damaged proteins <strong>and</strong> rapid turnover in response to<br />
physiological changes is performed by the ubiquitin system. Ubiquitin is aprotein that<br />
becomes attached via lysine residues in target proteins (Finley <strong>and</strong> Chau, 1991). Target<br />
proteins are attached to multiple molecules of ubiquitin in reactions catalysed by<br />
ubiquitin-protein ligase <strong>and</strong> ubiquitin-conjugating enzymes. The target protein-ubiquitin<br />
complex is recognized by amulti-enzyme proteolytic complex termed the proteosome.<br />
The latter catalyses the ATP-dependent degradation of the target protein to amino acids<br />
<strong>and</strong> peptides <strong>and</strong> releases ubiquitin so that it becomes available for further ubquinitation<br />
reactions.<br />
12.9 Yeast stress responses<br />
Yeast acidifies the medium when it is transferred from beer <strong>and</strong> suspended in water. This<br />
phenomenon is stimulated by the presence of exogenous glucose. The spontaneous <strong>and</strong><br />
glucose-induced acidification of media is used as a method of assessing yeast<br />
physiological condition (Sigler <strong>and</strong> Hofer, 1991). The extrusion of protons is<br />
accompanied by the excretion of several organic species, including amino acids <strong>and</strong><br />
nucleotides. The process has been termed shock excretion (Lewis <strong>and</strong> Phaff, 1964). It<br />
requires the presence of afermentable sugar <strong>and</strong> presumably represents astress-induced<br />
transient loss of membrane integrity.<br />
As yeast progresses through the brewing cycle of storage, pitching, fermentation,<br />
cropping <strong>and</strong> storage, it is subject to a number of stresses. These include rapid<br />
temperature fluctuation, high barometric pressure, high osmotic pressure, low water<br />
activity, low pH, high ethanol concentration, transient aerobiosis <strong>and</strong> starvation. Yeasts<br />
have evolved various metabolic strategies to minimize the deleterious effects of many<br />
stresses (Sorger, 1991). The ability to withst<strong>and</strong> an applied stress can be constitutive, for<br />
example, the differing degrees of ethanol tolerance observed in individual strains of yeast<br />
(Section 12.5.9). Similarly, different yeast species can grow over different temperature<br />
ranges, indicating a spectrum of thermotolerance. For example, ale strains of S. cerevisiae<br />
can usually grow at higher temperatures than lager strains. Resistance to some stresses<br />
can be induced by non-lethal exposure to the stress in question. The biochemistry of these<br />
acquired responses is distinct for individual stresses, although often there is overlap.<br />
Thus, exposure to one stress commonly induces tolerance to a range of other stresses.<br />
Strategies for overcoming many stresses involve common cellular mechanisms.<br />
The induced stress response requires a sensing system, a signal transduction pathway<br />
<strong>and</strong> expression of target genes. Exposure of yeast to elevated temperatures results in<br />
cellular dysfunction <strong>and</strong> ultimately death. These effects are caused via denaturation of<br />
proteins <strong>and</strong> membrane damage. When S. cerevisiae undergoes a temperature shift from<br />
25 ëC to 37 ëC growth is temporarily arrested, trehalose accumulates <strong>and</strong> the synthesis of<br />
most proteins ceases. A small group of around 70 so-called heat shock proteins (hsps) are<br />
induced <strong>and</strong> produced in high concentrations. After a short period, growth recommences<br />
<strong>and</strong> synthesis of hsps continues at a reduced but higher level than non-heat shocked yeast.<br />
The heat-shocked yeast exhibits elevated thermotolerance.<br />
The heat shock proteins include those responsible for trehalose biosynthesis. This is<br />
predictable, bearing in mind the membrane-stabilizing properties of this metabolite. The<br />
ubiquitin system is activated which facilitates the destruction of damaged proteins (12.8).<br />
Other hsps function as protein repair systems, ensuring that damaged molecules are<br />
refolded in their correct conformations. The mechanism by which the heat shock response<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
is generated has not been fully elucidated. The hsp genes are regulated by aheat shock<br />
transcription factor (hsf1P) that binds to the promoter region. The route by which the<br />
signal from heat shock to activation of the gene encoding hsf1P is not known. However,<br />
the protein can be phosphorylated <strong>and</strong> the extent of phosphorylation has been shown to<br />
correlate with transcriptional activation over arange of temperatures (Sorger, 1991).<br />
Other stress responses are mediated by aspecific stress response element (STRE)<br />
whichhasbeenidentifiedinthepromoterregionofseveralstressresponsivegenes.These<br />
include genesencodingforenzymesinvolved inprotectionagainst amultitude ofstresses<br />
including,oxidative, lowpH,ethanol,weakorganicacid<strong>and</strong>osmotic. Activationofthese<br />
stress genes, via the STRE element, is in response to at least two signal transduction<br />
pathways. These are the Ras cAMP mediated cascade (Section 12.5.8) <strong>and</strong> the high<br />
osmolarity glycerol (HOG) pathway (Section 12.3.1). The common regulatory element<br />
within anumber of apparently unconnected genes explains how individual stresses can<br />
induce overlapping responses.<br />
The relevance of stress responses in yeast to brewing is unclear. The multitude of<br />
stresses to which yeast is subject during fermentation imply that astress response will be<br />
exhibited throughout most if not all of the stages of brewing where yeast is present. It is<br />
reasonable to assume that astress response makes the yeast more resistant to the stresses<br />
of the brewing process. This may be of particular importance in the case of high-gravity,<br />
high-volume fermentations.<br />
12.10 Minor products of metabolism contributing to beer flavour<br />
Both ethanol <strong>and</strong> carbon dioxide contribute to beer flavour. The latter has a`mouth<br />
tingle' character, whereas ethanol imparts a `warming' note to beers. In addition,<br />
fermentation of wort generates amultitude of other minor products of yeast metabolism,<br />
many of which contribute to beer flavour. The action of yeast on wort also serves to<br />
removesomecomponentswhosepersistenceinbeerwouldbeundesirable.Theformation<br />
of adesirable mixture of flavour-active metabolites in beer is influenced by the choice of<br />
yeast strain <strong>and</strong> wort composition. One of the primary aims of fermentation management<br />
is to control the process to ensure that flavour metabolites are produced in consistent <strong>and</strong><br />
desired quantities.<br />
The principal flavour metabolites are aliphatic alcohols, aldehydes, organic <strong>and</strong> fatty<br />
acids <strong>and</strong> esters of alcohols <strong>and</strong> fatty acids. These are formed as by-products of the<br />
metabolism of sugars <strong>and</strong> amino acids. The relationships between these classes of<br />
metabolites are shown in Fig. 12.19. In addition amyriad of other products of yeast<br />
metabolism contribute to beer flavour. Many of these are excreted by yeast during<br />
fermentation. However, some are intracellular components that are released in the beer<br />
either by cell death <strong>and</strong> autolysis or via shock excretion (Section 12.9).<br />
12.10.1 Organic <strong>and</strong> fatty acids<br />
More than ahundred organic <strong>and</strong> fatty acids have been identified in yeast (Meilgard,<br />
1975). Although some of these are derived from wort, many are produced as aresult of<br />
yeast metabolism. Organic acid formation <strong>and</strong> excretion contributes to the reduction in<br />
pH that occurs during fermentation. They confer a`sour' or `salty' taste to beers. The<br />
most abundant organic acids found in beers <strong>and</strong> their typical concentrations are shown in<br />
Table 12.9. Organic acids are largely derived from the incomplete TCA cycle that occurs<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 12.9 Beer organic acids (Coote <strong>and</strong> Kirsop, 1974; Whiting, 1976; Klopper et al., 1986).<br />
Organic acid Typical concentration in beer (mg/l)<br />
Acetic 10±50<br />
Citric 100±150<br />
Lactic 50±300<br />
Malic 30±50<br />
-ketoglutaric 0±60<br />
Pyruvic 100±200<br />
Succinic 50±150<br />
Dicarboxylic acids<br />
Tricarboxylic acids<br />
Sugars Amino acids<br />
Acyl-CoA esters<br />
Esters<br />
Oxo-acids<br />
Aldehydes<br />
Alcohols<br />
Vicinal<br />
diketones<br />
Fig. 12.19 Relationships between the major classes of yeast-derived beer flavour compounds.<br />
during anaerobic repressed growth of yeast (Section 12.5.8). In addition, some may<br />
derive from the catabolism of amino acids (Fig. 12.18). Lactate is derived from the<br />
reduction of pyruvate. The extracellular concentrations of some organic acids may both<br />
increase <strong>and</strong> decrease during fermentation. For example, it has been reported that<br />
pyruvateisexcretedduringearly fermentation.Atalater stage,thisacidistakenupagain<br />
<strong>and</strong> acetate is excreted (Coote <strong>and</strong> Kirsop, 1974). The maximum exogenous pyruvate<br />
concentration coincides with the point at which wort free amino nitrogen ceases to be<br />
assimilated possibly suggesting that the former is derived from the dissimilation of the<br />
latter (Fig. 12.1). The accumulation of exogenous pyruvate indicates that at least during<br />
partoffermentation,therateofpyruvatedissimilationtoethanolisslowerthantherateof<br />
formation of pyruvate. The extracellular formation of two oxo-acids, -acetolactate <strong>and</strong><br />
-acetohydroxybutyrate is of special note since they the precursors of the vicinal<br />
diketones, diacetyl <strong>and</strong> 2,3-pentanedione (Section 12.10.2).<br />
Short <strong>and</strong> medium chain length fatty acids have unpleasant flavours <strong>and</strong> they inhibit<br />
beer foam formation. For these reasons, their presence in beer is undesirable. Generally,<br />
the medium chain-length fatty acids, principally C16 <strong>and</strong> C18, of wort are replaced by<br />
shorter chain-length fatty acids (C6±C10) in beer (Chen, 1980). These short chain-length<br />
fatty acids are powerful detergents <strong>and</strong> it seems probable that they are not excreted by<br />
yeast in acontrolled process. Instead, it is likely that they exit cells as aresult of plasma<br />
membrane leakage in response to ethanol stress (Section 12.5.9) or in extreme cases<br />
because of cell death <strong>and</strong> autolysis.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
12.10.2 Carbonyl compounds<br />
Carbonyl compounds are abundant in beers, where more than 200 have been detected<br />
(Berry <strong>and</strong> Watson, 1987). The concentrations of several aldehydes <strong>and</strong> the vicinal<br />
diketones are influenced by yeast metabolism during fermentation <strong>and</strong> subsequent<br />
conditioning. As agroup, these generally make anegative contribution to beer flavour<br />
<strong>and</strong> aroma. An important requirement of fermentation management is toensure that these<br />
compounds are reduced to acceptable concentrations.<br />
Several aldehydes arise during wort production, others are formed as intermediates in<br />
the biosynthesis of higher alcohols from oxo-acids by yeast (Fig. 12.19). Exogenous<br />
aldehydes form adducts with sulphur dioxide <strong>and</strong> in this form they may not be available<br />
for enzymatic reduction. In this sense, the metabolism by yeast of aldehydes <strong>and</strong> sulphur<br />
containing compounds are intimately related. Several yeast reductases each with a<br />
differing spectrum of activity are involved in the elimination of aldehydes (Debourg et<br />
al., 1993). These enzymes use either NADH or NADPH as hydrogen donor cofactors.<br />
The fermentative alcohol dehydrogenase (ADHI) is responsible for the reduction of<br />
pentanal <strong>and</strong> pentenal, <strong>and</strong> an aldose reductase reduces 3-methyl butanal <strong>and</strong> possibly<br />
pentanal. In addition, an aldoketoreductase with broad specificity has been detected<br />
(Laurent et al., 1995).<br />
Acetaldehyde is of special interest because of its role as the immediate precursor of<br />
ethanol. It has an unpleasant `grassy' flavour <strong>and</strong> aroma. Acetaldehyde is formed during<br />
the early to mid stages of fermentation <strong>and</strong> thereafter it declines to alow level. In some<br />
circumstances, it can accumulate during fermentation in concentrations above the flavour<br />
thresholdof10±20ppm.Theprincipal causes ofhighacetaldehyde concentrations inbeer<br />
are the use of poor quality pitching yeast, excessive wort oxygenation, unduly high<br />
fermentation temperature <strong>and</strong> excessive pitching rates (Geiger <strong>and</strong> Piendl, 1976).<br />
S. cerevisiae possesses two acetaldehyde dehydrogenases. One is mitochondrial <strong>and</strong><br />
requires NAD + orNADP + <strong>and</strong> K + for activity. The second enzyme is NADP + -linked,<br />
activated by Mg 2+ <strong>and</strong> is located in the cytosol. It has been proposed that the<br />
mitochondrial enzyme is functional only during oxidative growth on ethanol (Jacobsen<br />
<strong>and</strong> Bernofsky, 1974). However, in a study of the activities of both acetaldehyde<br />
dehydrogenases during ahigh-gravity lager fermentation this supposition was apparently<br />
disproved (Fig. 12.20). Here it may be seen that the activity of ADH1 correlated closely<br />
with the formation of ethanol. Surprisingly, the cytosolic Mg 2+ -dependent acetaldehyde<br />
dehydrogenase was active only during the early aerobic phase of fermentation. The K + -<br />
dependent mitochondrial acetaldehyde dehydrogenase was active throughout the whole<br />
of fermentation. Presumably, the latter enzyme would be instrumental in removing<br />
acetaldehyde during the later stages of fermentation. The reduction in acetaldehyde<br />
concentration characteristic of late fermentation would correlate with the concomitant<br />
increase in the concentration of extracellular acetate described in the previous section<br />
(Section 12.10.1).<br />
The concentrations of two vicinal diketones (VDK), diacetyl (2,3-butanedione) <strong>and</strong><br />
2,3-pentanedione are of critical importance in the fermentation of lager beers. Both<br />
compounds have strong `butterscotch' or `toffee' aromas <strong>and</strong> tastes. Their presence in<br />
lagers at concentrations higher than their flavour thresholds of around 0.15ppm <strong>and</strong><br />
0.9ppm respectively, causes an objectionable flavour defect. The now accepted pathway<br />
is that vicinal diketones arise as by-products of the synthesis of valine <strong>and</strong> isoleucine<br />
(Fig. 12.21). Aproportion of the pools of two acetohydroxy acids, -acetolactate <strong>and</strong> -<br />
acetohydroxybutyrate is excreted into the fermenting wort. There they undergo<br />
spontaneous oxidative decarboxylation to form diacetyl <strong>and</strong> 2, 3-pentanedione. In late<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 12.20 Specific activities of Mg 2+ -<strong>and</strong> K + aldehyde dehydrogenases (ALDH), alcohol<br />
dehydrogenase (ADH1) <strong>and</strong> changes in the concentrations of acetaldehyde <strong>and</strong> ethanol during a<br />
stirred 2.5 litre laboratory anaerobic fermentation using 12 ëP lager wort <strong>and</strong> a lager yeast strain<br />
(W. Tessier, unpublished data).<br />
fermentation, or during the conditioning phase, vicinal diketones are re-assimilated by<br />
yeast <strong>and</strong> reduced to form acetoin <strong>and</strong> 2,3-butanediol from diacetyl <strong>and</strong> 2,3-pentanediol<br />
from 2,3-pentanedione. The pattern of VDK formation <strong>and</strong> dissimilation during<br />
fermentation is shown in Fig. 12.1. The flavour thresholds of these reduced compounds<br />
are relatively high <strong>and</strong> at the concentrations that they are found in beer their presence is<br />
acceptable.<br />
The wort FAN concentration <strong>and</strong> amino acid spectrum influence the formation of<br />
acetohydroxy acids. Nakatani et al., (1984a,b) derived arelationship between total VDK<br />
concentration formed (T-VDKmax) <strong>and</strong> the minimum FAN concentration achieved during<br />
fermentation:<br />
T-VDKmax ˆ<br />
0:161<br />
FANmin 3:87 ‡0:415<br />
This relationship was taken to imply that wort composition <strong>and</strong> fermentation conditions<br />
should be manipulated to ensure that acontrolled residual FAN concentration was left at<br />
theendoffermentation. ThisprocedurewouldminimizethemagnitudeoftheVDKpeak.<br />
Similarly, worts with high concentrations of valine <strong>and</strong> isoleucine suppressed the<br />
formation of excessive VDK. Other than using worts with very high FAN concentrations,<br />
it is difficult to see how the spectrum of individual amino acids could be easily<br />
manipulated.<br />
The rate-determining step in the formation of VDK is the spontaneous oxidative<br />
decarboxylationoftheacetohydroxyacids.Thereactionsproceedrelativelyrapidlyunder<br />
aerobic conditions. Under anaerobic conditions, metal ions such as Cu 2+ ,Al 3+ <strong>and</strong> Fe 3+<br />
can act as alternative electron acceptors. The process is favoured by acidic conditions<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CH 3 CO CO CH 3<br />
Diacetyl<br />
Acetoin<br />
2,3-Butanediol<br />
CH3 CHOH CHOH CH3 CO2 + 2H<br />
NADH<br />
NAD +<br />
CH 3 CO COOH<br />
Pyruvate<br />
CH 3<br />
α-Acetolactate<br />
Thiaminepyrophosphateacetaldehyde<br />
TPP<br />
CH3 CO C<br />
OH<br />
COOH<br />
NADPH<br />
NADP +<br />
-NH2<br />
NADH<br />
NAD +<br />
2,3-Dihydroxy-isovalerate 2,3-Dihydroxy-3-methylvalerate<br />
H2O<br />
α-Keto-isovalerate α-Keto-3-methylvalerate<br />
-NH2<br />
CH3 CHOH CHNH2 COOH<br />
Threonine<br />
α-Ketobutyrate<br />
α-Acetohydroxybutyrate<br />
Valine Isoleucine<br />
CH3 CH2 CH3 CO C<br />
OH<br />
COOH<br />
CH 3 CHCH 3 CHNH 2 COOH CH 3 CH 2 CHCH 3 CHNH 2 COOH<br />
CO2 + 2H<br />
NADH<br />
NAD +<br />
CH 3<br />
CO CO CH2 OH<br />
2,3-Pentanedione<br />
2,3-Pentanediol<br />
CH3 CHOH CHOH CH2 CH3 Fig. 12.21 Pathways for the formation <strong>and</strong> dissimilation of the vicinal diketones, diacetyl <strong>and</strong> 2, 3-pentanedione.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(Inoue et al., 1968). Heating -acetolactate at 70ëC under anaerobic conditions results in<br />
non-oxidative decarboxylation directly to acetoin. It has been suggested that this can also<br />
occur at moderate temperatures providing that sufficiently low redox conditions are<br />
maintained (Inoue et al., 1991).<br />
Reduction of VDK occurs in late fermentation or during conditioning <strong>and</strong> it requires<br />
the presence of viable yeast. Under most circumstances, yeast assimilates <strong>and</strong> reduces<br />
free diacetyl very rapidly. Consequently, the VDK which can be detected in wort during<br />
fermentation is mostly precursor acetohydroxy acid. Before VDK analyses can be<br />
performed on samples removed during fermentation, they must first be heated to ensure<br />
that all the acetohydroxy acid is converted to VDK. The majority of brewery VDK<br />
analyses represent the sum of free diacetyl <strong>and</strong> -acetolactate. This is underst<strong>and</strong>able<br />
since -acetolactate is unstable <strong>and</strong> in the absence of yeast can be considered as<br />
`potential diacetyl'. An essential aspect of fermentation management is to ensure that<br />
yeast is not separated from green beer before the pools of free VDK <strong>and</strong> precursor<br />
acetohydroxy acids are reduced to an acceptable concentration.<br />
The physiological role fulfilled by VDK reduction during fermentation is not known,<br />
other than the possibility that it may represent another route for redox balancing (Section<br />
12.2). The enzymology of VDK reduction is not well characterized. Several yeast<br />
reductases, both NAD + <strong>and</strong> NADP + -requiring show activity in vitro towards diacetyl <strong>and</strong><br />
2,3-pentanedione, however, it is difficult to prove that they fulfil this role in vivo.<br />
Nevertheless, several reports of the occurrence of specific diacetyl reductases in S.<br />
cerevisiae have appeared in the literature (Louis-Eugene et al., 1988; Legay et al., 1989;<br />
Heidlas <strong>and</strong> Tressl, 1990; Scharwz <strong>and</strong> Hang, 1994; Murphy et al., 1996).<br />
In the last of these, it was shown that several brewing strains could be classified into<br />
two groups based on the differential patterns of thermolability of diacetyl reductases.<br />
Lager strains all contained a distinct heat stable acetoin reductase <strong>and</strong> alcohol<br />
dehydrogenases, which reduced diacetyl but showed no activity towards acetoin. Topfermenting<br />
ale strains lacked the heat stable acetoin reductase but all contained another<br />
enzyme, which would reduce both acetoin <strong>and</strong> diacetyl.<br />
The ability of yeast to reduce exogenous diacetyl declines during fermentation<br />
(Boulton <strong>and</strong> Box, 1999). This was demonstrated by adding exogenous free diacetyl to a<br />
series of similar stirred laboratory wort fermentations <strong>and</strong> monitoring the profiles of<br />
subsequent decline in concentration. The rates of uptake were slower, the later in<br />
fermentation that the diacetyl was added. At the end of fermentation when VDK<br />
concentrationswereclosetotheminimumspecificationrequiredbeforetheapplicationof<br />
cooling, yeast was most deficient in its ability to reduce VDK. This was taken to imply a<br />
possible decline in rates of diacetyl uptake because of lack of yeast membrane<br />
competence.<br />
12.10.3 Higher alcohols<br />
The formation of glycerol has been described (Section 12.3.1). Several alcohols, other<br />
than ethanol are formed in beer during fermentation (Engan, 1981). Higher alcohols<br />
achieve maximum concentrations in fermenting wort at atime roughly coincident with<br />
the point at which free amino nitrogen falls to aminimum concentration (Fig. 12.1).<br />
These are termed fusel alcohols because of their occurrence in fusel oil. This is a byproduct<br />
of the production of ethanol from the fermentation of carbohydrates. Those that<br />
contribute to beer flavour include n-propanol, iso-butanol, 2-methylbutanol <strong>and</strong> 3methlybutanol.<br />
It is considered that they impart a desirable warming character to beers<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
such that they intensify the flavour of ethanol. Higher alcohols are the precursors of the<br />
more flavour active esters.<br />
Some higher alcohols may derive from the reduction of aldehydes <strong>and</strong> ketones present<br />
in wort. Higher alcohols are synthesized from 2-oxo acids. These are decarboxylated to<br />
form the corresponding aldehyde <strong>and</strong> then reduced to the alcohol. The alcohol<br />
dehydrogenases are NAD + -dependent. This has prompted the suggestion that higher<br />
alcohol biosynthesis represents another mechanism for cellular redox control (Quain <strong>and</strong><br />
Duffield, 1985). The 2-oxo acids arise from carbohydrate metabolism via pyruvate <strong>and</strong><br />
acetyl-CoA, the so-called anabolic route, which forms part of the biosynthetic pathways<br />
of amino acids. Alternatively, the 2-oxo acid derives from atransamination reaction<br />
during amino acid utilization. This is termed the catabolic or Ehrlich route to higher<br />
alcohol formation (Fig. 12.22).<br />
The relative contribution of each biosynthetic route is determined by wort<br />
composition, the identity of the alcohol <strong>and</strong> the stage in fermentation. Thus, n-propanol<br />
is formed exclusively via the anabolic route since there is no corresponding amino acid.<br />
Predictably, where the wort has a high content of amino acids the catabolic route is<br />
favoured. Thus, under these conditions, amino acid synthesis is reduced via feed-back<br />
inhibition <strong>and</strong> the pool of 2-oxo acids is generated largely via amino acid catabolism. In<br />
the reverse situation where the supply of exogenous free amino acid is restricted, 2-oxo<br />
acids are formed via de novo synthesis from sugars <strong>and</strong> the anabolic route predominates.<br />
During fermentation of a typical all-malt wort, it has been reported that the yields of<br />
higher alcohols from each route are roughly similar (Schulthess <strong>and</strong> Ettlinger, 1978).<br />
During the early part of fermentation when free amino acids are relatively plentiful, the<br />
catabolic biosynthetic route predominates. This is gradually reversed as the concentration<br />
of assimilable amino nitrogen declines.<br />
Control of higher alcohol formation is achieved by the choice of an appropriate yeast<br />
strain <strong>and</strong> manipulation of fermentation conditions <strong>and</strong> wort composition. Several authors<br />
have reported that the choice of yeast strain has the biggest impact <strong>and</strong> that ale strains<br />
generally produce more higher alcohols than lager strains (Szlavko, 1974; Engan, 1978;<br />
Romano et al., 1992). Manipulation of fermentation conditions <strong>and</strong> wort composition in<br />
ways that favour increased yeast growth also tend to elevate higher alcohol concentration<br />
in beer. For example, high wort FAN <strong>and</strong> high wort dissolved oxygen. Use of high<br />
fermentation temperatures also favours increased levels of higher alcohols, possibly due<br />
to alterations in membrane fluidity (Peddie, 1990). Higher alcohol formation can be<br />
lowered by application of top pressure during fermentation (Rice et al.,1976).<br />
12.10.4 Esters<br />
Esters comprise the most important group of flavour-active compounds that are formed<br />
by yeast during fermentation. More than 100 esters have been detected in beers<br />
(Meilgard, 1975; Engan, 1981). Esters have fruity/solvent-like aromas <strong>and</strong> flavours. The<br />
most abundant is ethyl acetate, which accumulates to concentrations of 10±20 ppm. The<br />
concentrations of other esters are usually less than 1 ppm. The predominant route for<br />
formation is via the esterification of ethanol or a higher alcohol <strong>and</strong> a fatty acyl-CoA<br />
ester. Two enzymes are involved, an acyl-CoA synthetase <strong>and</strong> alcohol acyl transferase:<br />
R1COOH ‡ ATP ‡ CoASH ! R1COSCoA ‡ AMP ‡ PPI<br />
R1COSCoA ‡ R2OH ! R1COOR2 ‡ CoASH<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
-NH2<br />
TPP<br />
Acetaldehyde Acetaldehyde-TPP Pyruvate<br />
Leucine<br />
α-Ketoiso<br />
caproic acid<br />
3-Methyl<br />
butan-1-al<br />
3-Methyl<br />
butan-1-ol<br />
(CH3) 2 CH CH2 CH2OH NADH + H +<br />
NAD +<br />
Acetate<br />
NADH + H +<br />
CO2<br />
β-Isopropyl<br />
malic acid<br />
NADH + H +<br />
NAD +<br />
Acetyl-CoA<br />
-NH2<br />
CoA<br />
α-Isopropyl<br />
malic acid<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
CoA<br />
TPP<br />
α-Acetolactate<br />
Pyruvate<br />
2,3-Dihydroxyisovalerate<br />
α-Ketoisovalerate<br />
Isobutanal<br />
Valine Isobutanol<br />
(CH 3) 2<br />
CH<br />
CH 2OH<br />
CO2<br />
NAD +<br />
NH2<br />
CO2<br />
NADH + H +<br />
Fig. 12.22 Pathways for the formation of higher alcohols.<br />
Threonine<br />
NADP +<br />
CO2<br />
NADH + H +<br />
α-Ketobutyric acid Propanal<br />
CO2<br />
NAD +<br />
-NH2<br />
NAD +<br />
NADPH + H +<br />
α-Aceto-α-hydroxy<br />
butyric acid<br />
α,β-Dihydroxyβ-methylvaleric<br />
acid<br />
α-Keto-β-methyl<br />
valeric acid<br />
2-Methylbutan-1-al<br />
2-Methylbutan-1-ol<br />
CH3 CH2 CH(CH3) CH2OH n-Propanol<br />
Isoleucine<br />
NADH + H +<br />
CH 3 CH 2 CH 2OH
Acetyl-CoA <strong>and</strong> longer chain acyl-CoA esters can arise via the action of pyruvate<br />
dehydrogenase or acyl-CoA synthetase. During fermentation, the former route is<br />
unimportant (Section 12.5.5). The importance of the alcohol acyl transferase has been<br />
confirmed by the observation that mutants lacking the enzyme produce very low<br />
concentrations of esters (Lyness et al., 1997).<br />
The synthesis of esters requires the expenditure of metabolic energy suggesting that<br />
ester formation must fulfil an important metabolic role. It may be amechanism for<br />
regulating the ratio of acyl-CoA to free CoA (Thurston et al., 1981). Peak ester<br />
concentrations are reached after the formation of higher alcohols has ceased (Fig. 12.1).<br />
Rates of ester synthesis are maximal at the mid-point of fermentation coinciding with the<br />
cessation of lipid synthesis. Thus, when acetyl-CoA cannot be utilized by lipid synthesis,<br />
theformationofestersprovidesanalternativeuseforthissubstrate.Intermediatesoflipid<br />
biosynthesis influence ester formation. Supplementation of worts with the unsaturated<br />
fattyacid,linoleicacid(50mgl 1 )causesadramaticdecreaseinesterformation(Thurston<br />
et al., 1982). It was suggested that this effect was due to inhibition of alcohol acyltransferasebyunsaturatedfattyacids.Thiseffecthasbeenconfirmedbyothers(Yoshioka<br />
<strong>and</strong> Hashimoto, 1982a,b, 1984) <strong>and</strong> led to the proposal that ester <strong>and</strong> lipid syntheses are<br />
inverselycorrelated.Thisissupportedbytheobservationthatincreasingoxygensupplyto<br />
wort tends to decrease ester synthesis. In this case, oxygen promotes the synthesis of<br />
unsaturated fatty acids, which in turn reduces the activity of alcohol acyltransferase.<br />
It now appears that this effect is exerted at amore fundamental level (Malcorps et al.,<br />
1991; Fuji et al., 1997). These reports provided evidence that oxygen <strong>and</strong> linoleic acid<br />
causedrepressionofalcoholacyltransferase.Inlaterstudies(Dufour<strong>and</strong>Malcorps,1994)<br />
the same groups demonstrated the existence of multiple isozymes of alcohol<br />
acyltransferase. These have different substrate specificity <strong>and</strong> not all are subject to<br />
repression by oxygen <strong>and</strong> unsaturated fatty acids.<br />
The spectrum of esters produced during fermentation is controlled by the range <strong>and</strong><br />
substrate specificity of the alcohol acyltransferases possessed by individual yeast strains.<br />
The concentrations of esters produced by given yeast strains can be modulated by factors<br />
that influence the availability of acyl-CoA esters <strong>and</strong> lipid biosynthesis. In particular, the<br />
provision of oxygen is crucial.<br />
12.10.5 Sulphur-containing compounds<br />
Sulphur-containing compounds in beer produced via yeast metabolic activity arise from<br />
organic sulphur-containing compounds such as some amino acids <strong>and</strong> vitamins.<br />
Alternatively, they may be formed from inorganic wort constituents such as sulphate.<br />
Sulphate is transported into yeast via aspecific permease. Once in the cellit is reduced to<br />
sulphite in reactions that require ATP for energy. Thereafter, sulphite is reduced to<br />
sulphide via an NADP + -dependent reductase. The sulphide is then available for<br />
incorporation intoavarietyofsulphur-containing organicmetabolites(Fig.12.23).Under<br />
some circumstances appreciable levels of hydrogen sulphide accumulate in beer. The<br />
resultant sulphidic taste is an essential part of the flavour of some ales, for example<br />
Burton pale ales. Over-accumulation of hydrogen sulphide is undesirable. It can be<br />
controlled by ensuring that beers are exposed to copper, in the form of a piece of<br />
sacrificial pipe, which allows the formation of an insoluble sulphide. The pool size of Sadenosylmethionine<br />
has regulatory significance. Thus, its presence in the cell inhibits the<br />
transcription of all the genes, which encode the enzymes responsible for the uptake of<br />
sulphate, its reduction to sulphide <strong>and</strong> the synthesis of S-adenosylmethionine.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The presence of certain amino acids in wort modulates the metabolism of sulphurcontaining<br />
compounds (Gyllang et al., 1989). Supplementation of growth media with<br />
methionine increases the intracellular concentration of S-adenosylmethionine <strong>and</strong> causes<br />
the effects described already. Threonine in the medium reduces the activity of<br />
aspartokinase by feed-back inhibition (Fig. 12.23). This in turn reduces the pool sizes<br />
of O-acetylhomoserine <strong>and</strong> methionine. Hence, sulphite levels increase since the<br />
repressing effects of S-adenosylmethionine are relieved. Isoleucine also causes an<br />
increase in sulphite since its presence inhibits threonine utilization.<br />
Sulphite forms adducts with carbonyl compounds. This prompted the suggestion that<br />
carbohydrate metabolism may influence sulphur metabolism (Korch et al., 1991). These<br />
authors demonstrated acorrelation between wort glucose concentration <strong>and</strong> sulphite<br />
levels in beer. At high glucose levels, there was a concomitant increase in the<br />
concentrations of pyruvate <strong>and</strong> acetaldehyde. These carbonyls formed addition<br />
compounds with sulphite, thereby depriving the methionine synthetic pathway of<br />
sulphite <strong>and</strong> resulting in derepression of the same.<br />
The formation of sulphite during wort fermentation is influenced by the availability of<br />
amino acids (Dufour, 1991). During early fermentation, aplentiful supply of methionine<br />
<strong>and</strong> threonine causes repression of the sulphite synthetic pathway. In the phase of active<br />
fermentation, depletion of methionine <strong>and</strong> threonine derepresses the sulphite synthetic<br />
genes but sulphite does not accumulate because it is fully utilized for the synthesis of<br />
sulphur-containing amino acids. In mid to late fermentation yeast growth ceases, the<br />
amino acid pool is fully depleted <strong>and</strong> sulphite reductase activity declines to alow level.<br />
Under these circumstances, sulphite accumulation takes place.<br />
Accumulation of sulphite in beers during fermentation is desirable since it may form<br />
adducts with potential stalingcarbonylssuch astrans-2-nonenal.Inthisrespect,there isa<br />
positive correlation between sulphite levels <strong>and</strong> beer flavour stability. Individual<br />
carbonyls have varying affinities for sulphite adduct formation. It has been claimed that<br />
the rate of sulphite formation regulates the proportion of carbonyls bound as adducts <strong>and</strong><br />
those available for reduction by yeast (Dufour, 1991). It is essential for good flavour<br />
stability that sufficient sulphite is available to prevent the displacement of potential<br />
staling aldehydes from adducts by irreversible reactions with other beer components such<br />
as quinones <strong>and</strong> polyphenols.<br />
The sulphur-containing compound dimethyl sulphide (DMS) is an important flavour<br />
compound. At high concentrations, it has arelatively objectionable taste <strong>and</strong> aroma of<br />
cooked sweet corn. At moderate concentrations (30±100ppb) it is considered to be an<br />
essential componentoflagerbeers.TheprecursorofDMSisS-methylmethionine(SMM)<br />
which is acomponent of malt (Chapter 4). During the conversion of green malt to<br />
finished malt, SMM is converted to DMS <strong>and</strong> another related metabolite dimethyl<br />
sulphoxide (DMSO). The conditions employed during malting influence the proportions<br />
of DMS <strong>and</strong> DMSO formed. A kilning temperature greater than 60 ëC is required for<br />
appreciable DMSO formation (Parsons et al., 1977). SMM is converted to DMS by heat.<br />
The temperatures required for this conversion occur only during the malt <strong>and</strong> wort<br />
production stages of brewing. However, DMS is volatile <strong>and</strong> much is lost during mashing<br />
<strong>and</strong> wort boiling. DMSO is heat stable <strong>and</strong> persists unchanged through these stages.<br />
At collection into fermenter, wort contains a mixture of SMM, DMS <strong>and</strong> DMSO. The<br />
proportions of each depend upon the raw materials used for wort production <strong>and</strong> the<br />
conditions employed in its manufacture. Further conversion of SMM to DMS during<br />
fermentation is precluded by the low temperature. The residual SMM is assimilated by<br />
yeast <strong>and</strong> converted to methionine. Thus, the concentration of DMS in beer is determined<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
O-Acetyl serine<br />
ATP<br />
PPi<br />
NADPH<br />
ADP + NADP +<br />
NADPH<br />
NADP +<br />
Sulphate<br />
Adenosine 5'phosphosulphate<br />
(APS)<br />
ATP<br />
ADP<br />
3'-Phosphoadenosine-<br />
5'-phosphosulphate (PAPS)<br />
Sulphite<br />
Sulphide<br />
ATP sulphurylase<br />
APS kinase<br />
PAPS reductase<br />
Sulphite reductase<br />
Serine Cysteine Homocysteine<br />
Methionine<br />
S-Adenosylmethioine<br />
O-Acetylhomoserine<br />
Aspartokinase<br />
N2H COOH<br />
CH<br />
CH2 N<br />
C<br />
C<br />
CH2 HC<br />
N<br />
CH2 S CH2 O<br />
H H<br />
H H<br />
OH OH<br />
N<br />
CH<br />
C<br />
N<br />
Aspartic acid<br />
β-Aspartyl phosphate<br />
Aspartic<br />
β-semialdehyde<br />
Homoserine<br />
Homoserine<br />
Phosphoric acid<br />
α-Keto-β-methylvaleric<br />
acid<br />
α,β-Dihydroxyβ-methylvaleric<br />
acid<br />
α-Aceto-α-hydroxybutyric<br />
acid<br />
α-Ketobutyric acid<br />
Threonine Pyruvate<br />
Fig. 12.23 Pathways for the assimilation of inorganic sulphur compounds <strong>and</strong> their incorporation into sulphur-containing amino acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
largely during wort manufacture. However, yeast possesses a reductase activity, which<br />
allows the formation of DMS from DMSO (Gibson et al., 1985). The enzyme is<br />
apparently subject to nitrogen catabolite repression <strong>and</strong> is most active in late fermentation<br />
when the wort FAN concentration is minimal.<br />
12.11 References<br />
ALEXANDER, M. A. <strong>and</strong> JEFFRIES, T. W. (1990) Enz. Microb. Technol., 12, 2.<br />
ANDRE, B. (1995) Yeast, 11, 1575.<br />
ANDREASON, A. A. <strong>and</strong> STIER, T. J. B. (1953a) J. Cellul. Comparat. Physiol., 41, 23.<br />
ANDREASON, A. A. <strong>and</strong> STIER, T. J. B. (1953b) J. Cellul. Comparat. Physiol., 43, 271.<br />
ARVINDEKAR, A. U. <strong>and</strong> PATIL, N. B. (2002) Yeast, 19, 131.<br />
BARNETT, J. A. (1981) Adv. Carbohydrate Chem. Biochem. 39, 347.<br />
BARNETT, J. A., PAYNE, R. W. <strong>and</strong> YARROW, D. (1990) Yeasts, Characteristics <strong>and</strong> Identification, 2nd edn,<br />
Cambridge University Press, Cambridge.<br />
BERRY, D. R. <strong>and</strong> WATSON, D. C. (1987). `Production of organoleptic compounds'. In Yeast Biotechnology,<br />
D. R. Berry, I. Russell <strong>and</strong> G. G. Stewart eds, pp. 345±368. Allen & Unwin, London.<br />
BLACKWELL, K. J., SINGLETON, I. <strong>and</strong> TOBIN, J. M. (1995) Appl. Microb. Biotechnol., 43, 579.<br />
BOULTON, C. A. <strong>and</strong> QUAIN, D. E. (2001) <strong>Brewing</strong> Yeast <strong>and</strong> Fermentation, Blackwell Science, Oxford.<br />
BOULTON, C. A., BOX, W. G., QUAIN, D. E. <strong>and</strong> MOLZAHN, S. W. (1999) Proc. 27th Cong. Eur. Brew. Conv.,<br />
Cannes, 687.<br />
BOURDOT, S. <strong>and</strong> KARST, F. (1995) Gene, 165, 97.<br />
BRUINENBERG, P. M., VAN DIJKEN, J. P. <strong>and</strong> SCHEFFERS, W. A. (1983) J. Gen. Microbiol., 129, 953.<br />
BUSTERIA, J. R. <strong>and</strong> LAGUNAS, R. (1986) J. Gen. Microbiol, 132, 379.<br />
CARTWRIGHT, C. P., ROSE, A. H., CALDERBANK, J. <strong>and</strong> KEENAN, M. H. (1989) In The Yeasts, Vol. 3 (2nd edn)<br />
A. H. Rose <strong>and</strong> J. S. Harrison eds, pp. 5±56, Academic Press, London.<br />
CASEY, W. M., KEESLER, G. A. <strong>and</strong> PARKS, L. W. (1992) J. Bacteriol., 174, 7283.<br />
CHEN, E. C-H. (1980) J. Am. Soc. Brew. Chem., 38, 148.<br />
CIESSAROVA, V. J., SMOGROVICOVA, D. <strong>and</strong> DOMENY, Z. (1996) Folia Microbiologica, 41, 485.<br />
CLARKSON, S. P., LARGE, P. J., BOULTON, C. A. <strong>and</strong> BAMFORTH, C. W. (1991) Yeast, 7, 91.<br />
COOTE, N. <strong>and</strong> KIRSOP, B. H. (1974) J. Inst. Brew., 80, 474.<br />
COSTA, V., REISS, E., QUINTANILHA, A. <strong>and</strong> MORADAS-FERREIRA, P. (1993) Archiv. Biochem. Biophys., 300,<br />
608.<br />
CRABTREE, H. G. C. (1929) Biochem. J., 23, 536.<br />
CROWE, J. H., CROWE, L. M. <strong>and</strong> CHAPMAN, D. (1984) Science, 223, 701.<br />
CRUMPLEN, R., D'AMORE, T., SLAUGHTER, C. <strong>and</strong> RUSSELL, I. (1993) Proc. 24th Cong. Eur. Brew. Conv.,<br />
Oslo, 267.<br />
CRUMPLEN, R. M., SLAUGHTER, J. C. <strong>and</strong> STEWART, G. G. (1996) Lett. Appl. Microbiol., 23, 448.<br />
D'AMORE, T., PANCHAL, C. J. <strong>and</strong> STEWART, G. G. (1988) Appl. Environ. Microbiol., 54, 110.<br />
D'AMORE, T., PANCHAL, C. J., RUSSELL, I. <strong>and</strong> STEWART, G. G. (1990) Crit. Rev. Biotechnol., 9, 287.<br />
DAVELOOSE, M. (1987) MBAA Tech. Quart., 24, 109.<br />
DAWES, I. W. (1999) In The Metabolism <strong>and</strong> Molecular Physiology of Saccharomyces cerevisiae J. R.<br />
Richardson <strong>and</strong> M. Schweizer eds, pp. 277±326. Taylor & Francis, London.<br />
DE NOBEL, J. G., KLIS, F. M., RAM, A, VAN UNEN, H., PRIEM, J., MUNNIK, T. <strong>and</strong> VAN DEN ENDE, H. (1991)<br />
Yeast, 7, 589.<br />
DE WINDE, J. H. <strong>and</strong> GRIVELL, L. A. (1993) Prog. Nucleic Acid Res. <strong>and</strong> Molec. Biol., 46, 51.<br />
DEBOURG, A., LAURENT, M., DUPIRE, S. <strong>and</strong> MASSCHELEIN, C. A. (1993) Proc. 24th Cong. Eur. Brew.<br />
Conv., Oslo, 437.<br />
DOES, A. L. <strong>and</strong> BISSON, L. F. (1989) J. Bacteriol., 171, 1303.<br />
DOMBEK, K. M. <strong>and</strong> INGRAM, L. O. (1986) Appl. Environ. Microbiol., 52, 975.<br />
DUFOUR, J.-P. (1991) Proc. 23rd Conv. Eur. Brew., Conv, Lisbon, 209.<br />
DUFOUR, J.-P. <strong>and</strong> MALCORPS, P. (1994) Proc. 4th Aviemore Conf. Malting, Brew. <strong>and</strong> Dist., I. Campbell<br />
ed., pp. 137±145, IOB, London.<br />
ELEUTHERIO, E. C. A., DE ARUJO, P. S. <strong>and</strong> PANEK, A. D. (1993) Biochim. et Biophys. Acta, 1156, 263.<br />
ENGAN, S. (1978) Proc. 18th Cong. Eur. Brew. Conv., Zoeterwoede, 28.<br />
ENGAN, S. (1981) `Beer composition: volatile substances'. In <strong>Brewing</strong> Science, (J. R. A. Pollock, ed.), vol.<br />
2, pp. 98±105. Academic Press, London.<br />
FAERGEMAN, N. J., DIRUSSO, C. C., ELBERGER, A., KNUDSON, J. <strong>and</strong> BLACK, P. N. (1997) J. Biol. Chem., 272,<br />
8531.<br />
FELS, S., RECKELBUS, B. <strong>and</strong> GOSSELIN, Y. (1999) Proc. 7th IOB Conv., Africa Section, Nairobi, 147.<br />
FINLEY, D. <strong>and</strong> CHAU, V. (1991) Ann. Rev. Cell Biol., 7, 25.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
FRAENKEL, D. G. (1982) `Carbohydrate metabolism'. The Molecular Biology of the Yeast, Saccharomyces<br />
cerevisiae, Vol. 2, Metabolism <strong>and</strong> Biosynthesis, pp. 1±37. J. N. Strathern, E. W. Jones <strong>and</strong> J. R.<br />
Broach eds, Cold Spring Harbor Laboratory.<br />
FRANCOIS, J., BLAZQUEZ, M. A., ARINO, J. <strong>and</strong> GANCEDO, C. (1997) Yeast Sugar Metabolism, Biochemistry,<br />
Genetics, Biotechnology <strong>and</strong> Applications, F. K. Zimmermann <strong>and</strong> K.-D. Entian eds, pp. 285±311,<br />
Technomic Publishing Co., Lancaster, USA.<br />
FRIDOVICH, I. (1986) Archiv. Biochem. Biophys., 247, 1.<br />
FUJI, T., YOSHIMOTO, H., NAGASAWA, N. <strong>and</strong> TAMAI, T. (1996) J. Ferment. Bioeng., 81, 538.<br />
FUJI, T., KOBAYASHI, O., YOSHIMOTO, H., FURUKAWA, S. <strong>and</strong> TAMAI, Y. (1997) Appl. Environ. Microbiol.,<br />
63, 910.<br />
GAMO, F. J., MORENO, E. <strong>and</strong> LAGUNAS, R. (1995) Yeast, 11, 1393.<br />
GANCEDO, J. M. (1992) Eur. J. Biochem., 206, 297.<br />
GANCEDO, J. M. <strong>and</strong> GANCEDO, C. (1997) `Gluconeogenesis <strong>and</strong> catabolite inactivation'. Yeast Sugar<br />
Metabolism, Biochemistry, Genetics, Biotechnology <strong>and</strong> Applications, F. K. Zimmermann & K.-D.<br />
Entian eds, pp. 359±378, Technomic Publishing Co., Lancaster, USA.<br />
GANCEDO, C. <strong>and</strong> SERRANO, R. (1989) `Energy-yielding metabolism'. The Yeasts, vol. 3, 2nd edn, A. H.<br />
Rose <strong>and</strong> J. S. Harrison eds, pp. 205±259, Academic Press, London.<br />
GEIGER, E. <strong>and</strong> PIENDL, A. (1976) MBAA Tech. Quart., 13, 51.<br />
GIBSON, R. M., LARGE, P. J. <strong>and</strong> BAMFORTH, C. W. (1985) J. Inst. Brew., 91, 397.<br />
GONZALES, C. I. <strong>and</strong> MARTIN, C. E. (1996) J. Biol. Chem., 271, 25801.<br />
GRENSON, M. (1992) `Amino acid transporters: structure, function <strong>and</strong> regulation'. Molecular Aspects of<br />
Transport Proteins. J. De Pont, ed., pp. 219±245, Elsevier Science Publishers, London.<br />
GUERIN, B. (1991) `Mitochondria'. The Yeasts, Vol. 4, A. H. Rose <strong>and</strong> J. S. Harrison, eds, pp. 541±600,<br />
Academic Press, London.<br />
GUSTIN, M. C., MARTINAC, B., SAIMI, Y., CUTHBERTSON, M. R. <strong>and</strong> KUNG, C. (1986) Science, 233, 1995.<br />
GYLLANG, H., WINGE, M. <strong>and</strong> KORCH, C. (1989) Proc. 23rd Cong. Eur. Brew. Conv, Zurich, 347.<br />
HEIDLAS, J. <strong>and</strong> TRESSL, R. (1990) Eur. J. Biochem., 188, 165.<br />
HINNEBUSCH, A. G. (1997) CRC, Crit. Rev. Biochem., 21, 277.<br />
HISADA, K., SUZUKI, Y. <strong>and</strong> TAMURA, K. (2002) Trends in High Pressure Biosci. <strong>and</strong> Biotechnol., 5, 325.<br />
HORAK, J. (1986) Biochim. Biophys. Acta, 864, 223.<br />
HORAK, J. (1997) Biochim. Biophys. Acta, 1331, 41.<br />
INGLEDEW, W. M. <strong>and</strong> PATTERSON, C. A. (1999) J. Amer. Soc. Brew. Chem., 57, 9.<br />
INOUE, T., MASAYAMA, K., YAMAMOTO, Y. <strong>and</strong> OKADA, K. (1968) Reports of the Res. Lab. Kirin Brew.<br />
Co., 11, 9.<br />
INOUE, T., MURAYAMA, H., KAJINO, K., KAMIYA, T., MITSUI, T. <strong>and</strong> MAWATARI, M. (1991) Proc. 23rd Cong.<br />
Eur. Brew. Conv., Lisbon, 369.<br />
JACOBSEN, M. K. <strong>and</strong> BERNOFSKY, C. (1974) Biochim. Biophys. Acta, 350, 277.<br />
JACOBSEN, M. <strong>and</strong> THORNE, R. S. W. (1980) J. Inst. Brew., 86, 284.<br />
JAZWINSKI, S. M. (1999) Expt. Gerontol., 34, 1.<br />
JONES, R. P. (1987) Proc. Biochem., 22, 129.<br />
JONES, R. P. (1989) Enz. Microb. Technol., 9, 334.<br />
JONES, R. P. <strong>and</strong> GADD, G. M. (1990) Enz. Microb. Technol., 12, 402.<br />
JONES, R. P. <strong>and</strong> GREENFIELD, P. F. (1984) Proc. Biochem., April, 48.<br />
JONES, R. P. <strong>and</strong> GREENFIELD, P. F. (1987) Enz. Microb. Technol., 9, 334.<br />
JONES, M. <strong>and</strong> PIERCE, J. S. (1970) Proc. 25th Cong. Eur., Brew. Conv., Interlaken, 151.<br />
KIRSOP, B. H. (1974) J. Inst. Brew., 80, 252.<br />
KIRSOP, B. H. (1982) Topics in Enzyme <strong>and</strong> Ferment. Biochem., 6, 79.<br />
KLOPPER, W. J., ANGELINO, S. A. G. F., TUNING, B. <strong>and</strong> VERMEIRE, H.-A. (1986) J. Inst Brew., 92, 311.<br />
KOHNO, Y., EGAWA, Y., ITOH, S., NAGAOKA, S., TAKAHASHI, M. <strong>and</strong> MUKAI, K. (1995) Biochim. Biophys.<br />
Acta, 1256, 52.<br />
KOLAROV, J., KOLAROV, N. <strong>and</strong> NELSON, N. (1990) J. Biol. Chem., 265, 12711.<br />
KORCH, C., MOUNTAIN, H. A., GYLLANG, H., WINGE, M. <strong>and</strong> BREHMER, P. (1991) Proc. 23rd Conv. Eur.<br />
Brew. Conv., Lisbon, 201±208.<br />
KREGER-VAN RIJ, N. J. W. (1984) The Yeasts, a Taxonomic Study (3rd edn) Elsevier, Biomedical Press,<br />
Amsterdam.<br />
KREMS, B., CHARIZANIS, C. <strong>and</strong> ENTIAN, K.-D. (1995) Curr. Genetics, 27, 427.<br />
KRUCKENBERG, A. L. (1996) Arch. Microbiol., 166, 283.<br />
KURTZMAN, C. P. <strong>and</strong> FELL, J. W. (1998) The Yeasts, A Taxonomic Study (4th edn) C. P. Kurtzman <strong>and</strong> J.<br />
W. Fell, eds, Elsevier, Amsterdam.<br />
LAGUNAS, R. (1993) FEMS Microbiol. Rev., 104, 229.<br />
LAURENT, M., GELDORF, B., VAN NEDERVELDE, L., DUPIRE, S. <strong>and</strong> DEBOURG, A. (1995) Proc. 25th Cong.<br />
Eur. Brew. Conv., Brussels, 337.<br />
LEGAY, O., RATOMAHHENINA, R. <strong>and</strong> GALZY, P. (1989) Agric. Biol. Chem., 53, 531.<br />
LEWIS, M. J. <strong>and</strong> PHAFF, H. J. (1964) J. Bacteriol., 87, 1389.<br />
LILLIE, S. H. <strong>and</strong> PRINGLE, J. R. (1980) J. Bacteriol., 143, 1384.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
LOUIS-EUGENE, S., RATOMAHHENINA, R. <strong>and</strong> GALZY, P. (1988) Folia Microbiologica, 33, 38.<br />
LUYTEN, K., ALBERTYN, J., SKIBBE, W., PRIOR, B. A., RAMOS, J., THEVELEIN, J. M. <strong>and</strong> HOHMANN, S. (1995)<br />
EMBO J., 14, 1360.<br />
LYNESS, C. A., STEELE, G. M. <strong>and</strong> STEWART, G. G. (1997) J. Amer. Soc. Brew. Chem., 55, 141.<br />
MAEDA, T., WURGLER-MURPHY, S. M. <strong>and</strong> SATO, H. (1994) Nature, 369, 242.<br />
MAGER, W. H. <strong>and</strong> VARELA, J. C. S. (1993) Molec. Microbiol., 10, 253.<br />
MAJARA, M., O'CONNOR-COX, E. S. C. <strong>and</strong> AXCELL, B. C. (1996a) J. Amer. Soc. Brew. Chem., 54, 149.<br />
MAJARA, M., O'CONNOR-COX, E. S. C. <strong>and</strong> AXCELL, B. C. (1996b) J. Amer. Soc. Brew. Chem., 54, 221.<br />
MALCORPS, P., CHEVAL, J. M., JAMIL, S. <strong>and</strong> DUFOUR, J-.P. (1991) J. Amer. Soc. Brew. Chem., 49, 47.<br />
MANSURE, J. J. C., PANEK, A. D., CROWE, L. M. <strong>and</strong> CROWE, J. H. (1994) Biochim. Biophys. Acta, 1191, 309.<br />
MARDER, R., BECKER, J. M. <strong>and</strong> NAIDER, F. (1977) J. Bacteriol., 131, 906.<br />
MARECHAL, P. A., MARTINEZ DE MARANON, I., MOLIN, P. <strong>and</strong> GERVAIS, P. (1995) Internat. J. Food<br />
Microbiol., 28, 277.<br />
MEILGARD, M. (1975) MBAA Tech. Quart., 12, 151.<br />
MISHRA, P. (1993) `Tolerance of fungi to ethanol'. In Stress Tolerance of Fungi D. H. Jennings, ed,<br />
pp. 189±208, Marcel Dekker, New York.<br />
MITCHELL, P. (1979) Science, 206, 1148.<br />
MURPHY, C., LARGE, P. J., WADFORTH, C., DACK, S. J. AND BOULTON, C. A. (1996) Biotechnol. Appl.<br />
Biochem., 23, 19.<br />
NAGODAWITHANA, T. W. <strong>and</strong> STEINKRAUS, K. H. (1976) Appl. Environ. Microbiol., 31, 158.<br />
NAKATANI, K., TAKAHASHI, T., NAGAMI, K. <strong>and</strong> KUMADA, J. (1984a) MBAA Tech. Quart., 21, 73.<br />
NAKATANI, K., TAKAHASHI, T., NAGAMI, K. <strong>and</strong> KUMADA, J. (1984b) MBAA Tech. Quart., 21, 175.<br />
O'CONNOR-COX, E. S. C., LODOLO, E. J. <strong>and</strong> AXCELL, B. C. (1993) J. Amer. Soc. Brew. Chem., 51, 97.<br />
OLIVERA, H., GONZALEZ, A. <strong>and</strong> PENA, A. (1993) Yeast, 9, 1065.<br />
OURA, E., HAARASILTA, S. <strong>and</strong> LONDESBOROUGH, J. (1980) J. Gen. Microbiol., 118, 51.<br />
OZCAN, S. <strong>and</strong> JOHNSTON, M. (1999). Microbiol. Molec. Biol. Rev., 63, 554.<br />
OZCAN, S., VALLIER, L. G., FLICK, J. S., CARLSON, M. <strong>and</strong> JOHNSTON, M. (1997) Yeast, 13, 127.<br />
PARKS, L. W. (1978) CRC Crit. Rev. Microbiol., 6, 301.<br />
PARKS, L. W. <strong>and</strong> CASEY, W. M. (1995) Ann. Rev. Microbiol., 49, 95.<br />
PARSONS, R. P., WAINWRIGHT, T. <strong>and</strong> WHITE, F. H. (1977) Proc. 16th Conv. Eur. Brew. Conv., Amsterdam,<br />
115.<br />
PEDDIE, H. A. B. (1990) J. Inst. Brew., 96, 327.<br />
PHAFF, H. J., MILLER, M. W. <strong>and</strong> MRAK, E. M. (1978) The Life of Yeasts (2nd edn) Harvard University Press,<br />
USA.<br />
PIERCE, J. S. (1987) J. Inst. Brew., 93, 378.<br />
POSTMA, E., VERDUYN, C. SCHEFFERS, W. A. <strong>and</strong> VAN DIJKEN, P. J. (1989) Appl. Environ. Microbiol., 55,<br />
468.<br />
PRATT-MARSHALL, P. L., BREY, S. E., DE COSTA, S. D., BRYCE, J. H. <strong>and</strong> STEWART, G. G. (2002) Brewers'<br />
Guardian, 131, 22.<br />
QUAIN, D. E. <strong>and</strong> DUFFIELD, M. L. (1985) Proc. 20th Cong. Eur. Brew. Cong., Helsinki, 307.<br />
QUAIN, D. E. <strong>and</strong> TUBB, R. S. (1982) MBAA Tech. Quart., 19, 29.<br />
RAPOPORT, A. I., KHRUSTALEVA, G. M., CHAMMANIS, G. Y. <strong>and</strong> BEKER, M. E. (1995) Mikrobiologiya, 64,<br />
229.<br />
RATTRAY, J. B. M. (1988) `Yeasts' In Microbial Lipids, C. Ratledge <strong>and</strong> S. G. Wilkinson eds, vol. 1,<br />
pp. 555±697, Academic Press, London.<br />
REES, E. M. R. <strong>and</strong> STEWART, G. G. (1997) J. Inst. Brew., 103, 287.<br />
RICE, J. F., CHICOYE, E., HELBERT, J. R. <strong>and</strong> GARVER, J. (1976) J. Amer. Soc. Brew. Chem., 35, 35.<br />
ROMANO, P., SUZZI, G., COMI, G. <strong>and</strong> ZIRONI, R. (1992) J. Appl. Bacteriol., 73, 126.<br />
ROSE, A. H. (1987) `Responses to the chemical environment'. In The Yeasts, 2nd edn, A. H. Rose <strong>and</strong> J. S.<br />
Harrison, eds, Vol. 2, pp. 5±40. Academic Press, London.<br />
ROSE, A. H. <strong>and</strong> HARRISON, J. S. (1995) `Responses to the chemical environment.' In The Yeasts (2nd edn)<br />
Vol. 2, Academic Press, London.<br />
RYDER, D. S., MURRAY, J. P. <strong>and</strong> STEWART, M. (1978) MBAA Tech. Quart., 15, 9.<br />
SCHNEIDER, R., BRORS, B., BURGER, F., CAMRATH, S. <strong>and</strong> WEISS, H. (1997) Curr. Genetics, 32, 384.<br />
SCHULTHESS, D. <strong>and</strong> ETTLINGER, L. (1978) J. Inst. Brew., 84, 240.<br />
SCHWARZ, J. G. <strong>and</strong> HANG, Y. D. (1994) Lett. Appl. Microbiol., 18, 272.<br />
SCHWEIZER, M. (1999) `Lipids <strong>and</strong> membranes'. In The Metabolism <strong>and</strong> Molecular Physiology of<br />
Saccharomyces cerevisiae, J. R. Richardson <strong>and</strong> M. Schweizer eds, pp. 79±155. Taylor & Francis,<br />
London.<br />
SERRANO, R. (1989) Ann. Rev. Plant Physiol. <strong>and</strong> Plant Molec. Biol., 40, 61.<br />
SIGLER, K. <strong>and</strong> HOFER, M. (1991) Biochim. Biophys. Acta, 1071, 375.<br />
SINGH, K. S. <strong>and</strong> NORTON, R. S. (1991) Arch. Microbiol., 156, 38.<br />
SORGER, P. K. (1991) Cell, 65, 363.<br />
SOUMALAINEN, H. <strong>and</strong> OURA, E. `Yeast nutrition <strong>and</strong> solute uptake'. (1971) In The Yeasts, A. H. Rose <strong>and</strong><br />
J. S. Harrison, eds, Vol. 2, pp. 3±74. Academic Press, London.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
STANLEY, G. A. <strong>and</strong> PAMENT, N. B. (1993) Biotechnol. Bioeng., 42, 24.<br />
STARK, M. J. (1999) `Protein phosphorylation <strong>and</strong> dephosphorylation' In The Metabolism <strong>and</strong> Molecular<br />
Physiology of Saccharomyces cerevisiae J. R. Richardson <strong>and</strong> M. Schweizer eds, pp. 209±275.<br />
Taylor & Francis, London.<br />
STEWART, G. G., ZHENG, X. <strong>and</strong> RUSSELL, I. (1995) Proc. 25th Cong. of the Eur. Brew. Conv., Brussels,<br />
403.<br />
SZLAVKO, C. M. (1974) J. Inst. Brew., 80, 534.<br />
THEVELEIN, J. M. (1994) Yeast, 10, 1753.<br />
THEVELEIN, J. M. <strong>and</strong> HOHMANN, S. (1995) Trends in Biochem. Sci., 20, 3.<br />
THURSTON, P. A., QUAIN, D. E. <strong>and</strong> TUBB, R. S. (1981) Proc. 18th Cong. Eur. Brew. Conv., Copenhagen,<br />
197.<br />
THURSTON, P. A., QUAIN, D. E. <strong>and</strong> TUBB, R. S. (1982) J. Inst. Brew., 88, 90.<br />
TUITE, M. F. (1991) `Protein synthesis'. In The Yeasts, Vol. 3, 2nd edn, A. H. Rose <strong>and</strong> J. S. Harrison eds,<br />
pp. 161±204, Academic Press, London.<br />
TZAGOLOFF, A. <strong>and</strong> DIECKMANN, C. I. (1990) Microbiol. Rev., 54, 211.<br />
VAN DER REST, M., KAMMINGA, A. H., NAKANO, A., ANRAKU, Y. <strong>and</strong> POOLMAN, W. N. (1995) Microbiol.<br />
Rev., 59, 304.<br />
VAN ZYL, P. J., KILIAN, S. G. <strong>and</strong> PRIOR, B. A. (1990) Appl. Microbiol. Biotechnol., 34, 231.<br />
VISSER, W., SCHEFFERS, W. A., BATENBURG VAN DER VEGTE, W. <strong>and</strong> VAN DIJKEN, J. P. (1990) Appl.<br />
Environ. Microbiol., 56, 3785.<br />
WALKER, G. M. (1998) Yeast Physiology <strong>and</strong> Biotechnology, John Wiley & Sons, West Sussex, UK.<br />
WALKER, G. M., BIRCH, R. M., CHANDRASENA, G. <strong>and</strong> MAYNARD, A. I. (1996) J. Amer., Soc. Brew. Chem.,<br />
54, 13.<br />
WANG, S. S. <strong>and</strong> BRANDRISS, M. C. (1987) Molec. Cell. Biochem., 7, 4431.<br />
WARBURG, O. (1926) Biochemische Zeitschrift, 172, 432.<br />
WHITING, G. C. (1976) J. Inst. Brew., 82, 84.<br />
WIAME, J.-M., GRENSON, M. <strong>and</strong> ARST, H. N. (1985) Adv. Microb. Physiol., 26, 1.<br />
WILKINSON, B. M., JAMES, C. M. <strong>and</strong> WALMSLEY, R. M. (1996) Microbiology, 142, 1667.<br />
YOSHIOKA, K. <strong>and</strong> HASHIMOTO, N. (1982a) Reports of Res. Lab. Kirin <strong>Brewing</strong> Co., 25, 1.<br />
YOSHIOKA, K. <strong>and</strong> HASHIMOTO, N. (1982b) Reports of Res. Lab. Kirin <strong>Brewing</strong> Co., 25, 7.<br />
YOSHIOKA, K. <strong>and</strong> HASHIMOTO, N. (1984) Reports of Res. Lab. Kirin <strong>Brewing</strong> Co., 27, 23.<br />
YOUATT, J. (1993) Crit. Rev. Microbiol., 19, 83.<br />
ZITOMER, R. S. <strong>and</strong> LOWRY, C. V. (1992) Microbiol. Rev., 56, 1.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
13<br />
Yeast growth<br />
13.1 Introduction<br />
Yeast growthis defined as the coordinated assimilation of nutrients from the medium<strong>and</strong><br />
subsequent metabolism to yield new biomass. New biomass is generated by increase in<br />
size of individual cells <strong>and</strong> by cellular proliferation. The biochemical reactions which<br />
underpin anabolic metabolism <strong>and</strong> which result in the synthesis of cellular<br />
macromolecules are outlined in Chapter 12. The biology of the yeast cell cycle is<br />
described in Chapter 11. In this chapter the dynamics of yeast populations with respect to<br />
the influence of cultural conditions are discussed.<br />
All organisms must proliferate <strong>and</strong> in so doing promulgate their genotypes via their<br />
progeny. The formation of beer during the fermentation of wort is aby-product of<br />
yeast growth. The aim of the brewer is to manipulate conditions to control the growth<br />
<strong>and</strong> metabolism of yeast to produce a desired product. In <strong>practice</strong>, this involves<br />
exerting appropriate controls to influence the balance between the yields of biomass<br />
<strong>and</strong> metabolites. With regard to batch fermentation, maximum fermentation efficiency<br />
isachievedbyminimizingtheproportionofwortnutrientsusedforbiomassgeneration<br />
<strong>and</strong> thereby maximizing the yield of beer. With regard to fermenter cycle times, a<br />
secondary aim is to ensure minimum residence times. Thus, fermentation management<br />
requires control of both yeast proliferation <strong>and</strong> growth rate. However, both of these<br />
aims have to be tempered by the need to employ conditions that yield beer of the<br />
desired quality.<br />
Most brewers ensure the trueness-to-type of production yeast strains by the periodic<br />
introduction of a new culture derived from a laboratory stock. Sufficient yeast for<br />
brewing is obtained by performing aseries of fermentations of ever-increasing volume.<br />
Since the aim is to generate yeast mass <strong>and</strong> not beer, the conditions are manipulated to<br />
favour growth. Conversely, in recent years there has been arenaissance of interest in<br />
continuous fermentation processes, especially using immobilized yeast. These may be for<br />
primary fermentation or perhaps more commonly for asecondary conditioning process<br />
such as continuous diacetyl removal. The aim of all of these processes is to minimize cell<br />
proliferation <strong>and</strong> use the yeast as a biocatalyst (Boulton <strong>and</strong> Quain, 2001).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Most modern brewers use pure cultures of asingle or defined mixture of yeast strains.<br />
This requires the resources of alaboratory, where stock cultures must be maintained. The<br />
appropriate apparatus must be available for growing pure cultures to produce sufficient<br />
yeast to inoculate a brewery propagation vessel. Methods are required to identify<br />
individual strains to provide assurance that the correct one is being used <strong>and</strong> that it is not<br />
contaminated with others. In addition, it is necessary to evaluate stored pitching yeast. At<br />
its simplest, this may take the form of ameasurement of viability. Increasingly, tests are<br />
being developed which assess the physiological condition of the viable fraction.<br />
13.2 Measurement of yeast biomass<br />
Yeast biomass can be measured in several ways (Table 13.1). Individual methods have<br />
advantages <strong>and</strong> disadvantages. No single procedure is absolutely precise <strong>and</strong> the results<br />
of individual methods are not always comparable. Care must be taken, therefore with the<br />
interpretation of results. This confusing situation is explicable in that different methods<br />
have been developed to fulfil specific needs. Typically, acompromise is made between<br />
rapidity <strong>and</strong> precision. Thus, although all methods have some shortcomings, individual<br />
procedures are useful providing they are used for the application that they were designed<br />
for.<br />
Direct methods of biomass determination use two general approaches. The first group<br />
of methods are those which measure the proportion of weight (or volume) due to cells<br />
within a suspension. The simplest method is to centrifuge a sample of slurry in a<br />
graduated tube. The volume fraction of packed yeast cells can be read directly from the<br />
scale on the tube. More commonly, the cell fraction is recovered by centrifugation or<br />
filtration from aslurry sample of known weight. The biomass concentration is expressed<br />
aspercentage wetweight ofyeastsolids. Both proceduresare rapidbutsubjecttoanerror<br />
if trub is present. This can be minimized by treating with alkali, which dissolves some of<br />
the non-yeast solid material.<br />
Assessment of yeast biomass concentration based on wet weight introduces an error<br />
due to variable amounts of the liquid phase, which are trapped in the interstices of the<br />
packed cells. This can be overcome by taking asample of slurry of known weight or<br />
volume, washing the cells in water to remove the suspending medium <strong>and</strong> drying to<br />
remove both intra <strong>and</strong> extracellular liquid. Biomass concentration is expressed as dry<br />
weight per unit volume of slurry. This method provides ahigh degree of precision, but it<br />
is time-consuming. The second direct approach is to count the numbers of cells<br />
suspended in aliquid using amicroscope <strong>and</strong> ahaemocytometer counting chamber<br />
(Section 13.10). Competently performed the procedure is precise <strong>and</strong> repeatable. It is<br />
rapid <strong>and</strong> therefore suitable where analyses are required for calculation or checking of<br />
pitching rates, etc. Since the yeast is examined directly, there is an opportunity to identify<br />
abnormalities or gross contamination. When used in conjunction with a vital stain the<br />
proportions of viable <strong>and</strong> dead cells can be estimated. The result is not affected by trub,<br />
but errors accrue where the yeast is heavily flocculent or a chain-former. The operator is<br />
the principal source of error.<br />
The error associated with visual examination of yeast is eliminated by the use of<br />
electronic particle counters. These devices rely on suspending yeast in an electrolyte <strong>and</strong><br />
passing the cells through a narrow orifice. Cells are registered in response to a change in<br />
electrical impedance. Most instruments can discriminate between particles of different<br />
sizes, reducing the error due to non-yeast solids. Nevertheless, any particle of similar size<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
to yeast cells registers in the result. Small yeast flocs or chains of cells introduce a further<br />
source of error. Problems associated with flocs can be reduced by prior treatment with a<br />
deflocculating agent such as maltose. This method is rapid <strong>and</strong> is most suitable for<br />
checking relatively low yeast counts such as are encountered in newly pitched worts or<br />
cask beers at rack.<br />
The classical technique for determining cell concentration is the plate count. A sample<br />
of the test suspension is serially diluted <strong>and</strong> aliquots are spread onto plates of selected<br />
solidified medium. Plates are incubated under suitable conditions for the test organism<br />
<strong>and</strong> growth is allowed to proceed until discrete colonies are formed. These are counted<br />
<strong>and</strong> it is assumed that each arises from a single cell, therefore the colony count is directly<br />
related to the cell concentration in the test suspension. This approach has several<br />
advantages. It is possible to use selective media so that mixed populations of cells can be<br />
identified. It only detects viable organisms. The precision of the method is low since not<br />
all viable cells grow to form colonies <strong>and</strong> flocs or chains do not develop into discrete<br />
colonies. The method is slow, although it is possible to use a rapid slide culture technique<br />
where cells are detected as micro-colonies. This reduces incubation times from a few<br />
days to a few hours. This method provides historical data only <strong>and</strong> is mainly used for<br />
checking strain purity <strong>and</strong> for detecting contaminants.<br />
Other indirect methods of biomass measurement are calibrated against one of the two<br />
direct approaches. Thus, an empirical relationship is established between biomass<br />
concentration <strong>and</strong> the measured parameter. Most of these approaches have no value in<br />
brewing. Two methods are used in brewing because they employ sensors which provide<br />
automatic in-line measurement of biomass concentration. The first detects yeast cells by<br />
light scattering. The second relies on the dielectric properties of intact cell membranes.<br />
The optical device uses a sensor that detects yeast in response to near infra-red<br />
radiation (Reiss, 1986). The control system doses yeast into wort to achieve a predetermined<br />
set-point of light scattering. This set-point is empirically derived for each<br />
yeast strain. A dual beam arrangement corrects for light scattering due to non-yeast solids<br />
in the unpitched wort. The shortcomings of this approach are that it does not correct for<br />
non-yeast solids present in the yeast slurry, it cannot be used with very flocculent strains<br />
<strong>and</strong> it requires a separate viability correction.<br />
From an electrical st<strong>and</strong>point, yeast cells suspended in beer or wort comprise a<br />
conducting medium <strong>and</strong> conducting cytoplasm, separated by a non-conducting plasma<br />
membrane. This juxtaposition allows cells to function as capacitors when subject to<br />
radiation of a suitable wavelength. In the case of yeast cells this is in the range of radio<br />
waves (Harris et al., 1987). The measured capacitance is proportional to the total volume<br />
fraction bounded by membrane within the operating field. Since yeast cells of a given<br />
strain, grown under defined conditions, are of a relatively constant size, measured<br />
capacitance is directly proportional to yeast biomass concentration. A biomass probe<br />
based on this principle is used in production brewing. The measured capacitance can be<br />
calibrated against cell number or a derivative of cell mass. A separate calibration curve is<br />
required for each strain.<br />
It has been successfully applied in systems for the automatic control of both yeast<br />
pitching <strong>and</strong> cropping (Boulton et al., 1989; Boulton <strong>and</strong> Clutterbuck, 1993). This<br />
method has several advantages. It has a very wide operating range <strong>and</strong> can be used to<br />
quantify cell concentrations in pitching yeast slurries without the need for dilution.<br />
Providing the cell suspension is homogenous, the concentrations of flocculent, nonflocculent<br />
<strong>and</strong> chain formers can be determined with equal facility. Non-yeast solid<br />
materials do not interfere since they do not function as capacitors. For the same reason,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 13.1 Methods for determining biomass concentration<br />
Method Timing Comments<br />
Direct methods:<br />
1. Determination of % wet<br />
weight of yeast in slurry<br />
2. Determination of % dry<br />
weight of yeast in slurry<br />
3. Cell count with<br />
haemocytometer <strong>and</strong><br />
microscope<br />
4. Electronic particle counters<br />
Rapid<br />
Slow<br />
Rapid<br />
Rapid<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Suitable for direct weight of pressed yeast cake<br />
Suitableforanalysisofpitchingyeastslurriesbycentrifugationorfiltration(IOB,1997(Section1.15.1,p.9))<br />
Errors due to entrained trub<br />
Not suitable for low yeast counts (cask beer counts, pitched worts)<br />
Reference method but requires 3 days at 110 ëC for reliable answer<br />
Errors due to entrained trub<br />
Suitable for all production analyses but requires skilled operator (IOB, 1997 (Section 1.15.1, p. 9))<br />
No error due to trub<br />
Not suitable for very flocculent strains<br />
Provides opportunity to view yeast <strong>and</strong> detect abnormalities or gross contamination<br />
Can be used to determine viability in conjunction with vital staining<br />
Suitable for quantifying yeast counts at low concentrations (cask beers, pitched worts etc.) (IOB, 1997<br />
(Section 1.15.1, p. 9))<br />
Not suitable for very flocculent <strong>and</strong> chain-forming strains<br />
Possible errors due to entrained trub
Indirect methods:<br />
1. Plate count Slow Cells inoculated onto solid medium <strong>and</strong> after incubation colonies are counted. It is assumed that each cell<br />
gives rise to a separate colony (IOB Methods of Analysis, 1997)<br />
Not all cells in original suspension may be cultivable<br />
Errors with flocculent cells<br />
Can be made selective by choice of suitable medium<br />
2. Nephelometry Rapid Light scattering at visible wavelengths used for rapid measurement of cell concentrations in laboratory<br />
research studies. Must be calibrated against other reference method since response is non-linear<br />
NIR light scattering used in a commercial automatic in-line pitching rate control system (Reiss, 1986)<br />
3. Analysis of cellular<br />
constituents<br />
Slow Biomass concentration expressed relative to the concentration of a macromolecular constituent such as DNA<br />
or protein<br />
Only useful as laboratory research tool<br />
4. Metabolic activity Rapid Biomass concentration inferred from a measure of metabolic activity such as the rate of oxygen uptake, CO2<br />
generation, exothermy<br />
Only used as a research tool<br />
5. Radiofrequency permittivity Rapid Biomass probe which infers cell concentration from capacitance measured in response to radiofrequency field<br />
(Harris et al., 1987)<br />
Requires separate calibration for each yeast strain<br />
Suitable for laboratory <strong>and</strong> in-line use<br />
Responsive only to viable fraction of yeast slurries<br />
Automatic in-line systems for control of cropping <strong>and</strong> pitching (Boulton et al., 1989; Boulton <strong>and</strong><br />
Clutterbuck, 1993)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
theprobecanbeusedtoquantifyyeastmixed withinertsupportmaterialsinimmobilized<br />
cell reactors. Yeast cells with disrupted plasma membranes, <strong>and</strong> which score as being<br />
non-viable with avital stain such as methylene blue, lose their dielectric properties <strong>and</strong><br />
are not detected. Thus, the probe is responsive only to viable cells.<br />
For any given yeast strain, grown under defined conditions, cell number <strong>and</strong> biomass<br />
weight are positively correlated. However, the relationship is dependent on the<br />
physiological state of the cell. For example, prolonged storage of pitching yeast results<br />
in the depletion of intracellular glycogen. In brewing yeast this reserve material can<br />
account for up to 30% of the cell dry weight (Section 12.5.7). Thus, in such aslurry, a<br />
plot of cell numbers would show little change with time until apoint is reached when<br />
cells die <strong>and</strong> lysis occurs. However, asimilar plot of cell dry weight would show a<br />
gradual decrease throughout the storage phase as glycogen reserves were dissimilated.<br />
13.3 Batch culture<br />
In abatch culture cells are inoculated into amedium with afinite supply of nutrients.<br />
This is the most common situation in commercial fermentations, including brewing. It is<br />
the most usual mode of growth in natural habitats. Anumber of distinct phases can be<br />
recognized (Fig. 13.1). These phases are common to all micro-organisms that reproduce<br />
by cell division. However, the budding habit of Saccharomyces yeast introduces some<br />
specific features. The duration of each phase is dependent on cultural conditions, the<br />
nature of the growth medium <strong>and</strong> the physiological condition of the inoculum. The lag<br />
phase represents aperiod of adaptation when cells undergo atransition from one set of<br />
conditions to another. Thus, cells in the inoculum may have to synthesize the enzymes<br />
that are needed for the uptake <strong>and</strong> utilization of substrates present in the medium. In<br />
addition, abrupt changes in temperature or osmotic potential may induce a stress<br />
response, which requires aperiod of recovery from before growth commences.<br />
Thecellspresent inaninoculumarenotidentical.Inthecase ofbuddingyeast,suchasS.<br />
cerevisiae, there may be differences in physiological condition, stage in the cell cycle <strong>and</strong><br />
cellular age. The duration of the lag phases of individual cells varies, <strong>and</strong> afinite period is<br />
required before the metabolic changes associated with adaptation are completed. The result<br />
is a progressive transition from lag phase to growth phase <strong>and</strong> hence, the period of<br />
accelerating growth. The growth rate gradually increases until it reaches amaximum <strong>and</strong><br />
constant value. This is the exponential growth phase where the population increases<br />
logarithmically. The nature of the organism, the composition of the medium <strong>and</strong> physical<br />
conditions such as temperature, pressure <strong>and</strong> degree of agitation determine the rate. Cells<br />
removed at this stage <strong>and</strong> transferred to afresh batch of similar medium will continue to<br />
grow exponentially without alag phase. Individual cells within the exponentially growing<br />
population divide at different times, but the graph of biomass against time increases<br />
smoothly since it represents the averaged concentration for the whole population. Such<br />
cultures are termed asynchronous. Various experimental procedures can be followed to<br />
synchronize the cell cycles of the entire population. For example, treatment with amitotic<br />
inhibitor causes cells to arrest at the same stage in the cell cycle. When the inhibitor is<br />
removed all the cells progress through the cell cycle in synchrony. Synchrony is usually<br />
maintained for the first two or three rounds of budding <strong>and</strong> plots of biomass concentration<br />
against time are stepped. After this time the culture gradually returns to the usual<br />
unsynchronized mode. In the case of budding yeast the loss of synchrony is hastened by the<br />
differingcellcycletimesofmother<strong>and</strong>daughtercells(Section11.7).Theexponentialphase<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Cell concentration<br />
1 2 3 4 5 6 7<br />
Time<br />
Fig. 13.1 Growth of yeast in batch culture on amedium such as wort in which sugar is the<br />
principal carbon source. For adetailed explanation see the text. The phases indicated are: 1, lag<br />
phase; 2, period of accelerating growth; 3, exponential growth phase; 4, decelerating growth phase;<br />
5, stationary phase; 6, diauxic shift phase; 7, second growth phase on ethanol if oxygen present<br />
(solid line), or death phase if no oxygen present (dotted line).<br />
of growth may end when a by-product of metabolism increases to an inhibitory<br />
concentration. More usually it terminates when an essential component of the medium<br />
becomes depleted. Eventually, the concentration of the essential nutrient falls to alevel at<br />
which it limits the growth rate <strong>and</strong> the culture moves into the decelerating phase of growth.<br />
Whengrowthceasesthecellsenterthestationaryphase.Duringthisphasethebiomass<br />
remains at aconstant level. This apparent inactivity is illusory since metabolic activities<br />
of various kinds continue. Aerobic cultures of S. cerevisiae growing on glucose as the<br />
principalcarbonsourcehaveabi-phasicgrowthcurve,termeddiauxie(Fig.13.1).During<br />
the first phase, cells grow at the expense of glucose. The effects of catabolite repression<br />
(Section 12.5.5) ensure that metabolism is fermentative <strong>and</strong> ethanol is produced. During<br />
the first stationary phase, the absence of glucose relieves catabolite repression. After a<br />
period of adaptation <strong>and</strong> providing oxygen is available the derepressed cells are able to<br />
utilize the ethanol <strong>and</strong> asecond phase of growth occurs.<br />
When no further growth is possible the cells remain in the stationary phase, often<br />
referred to as the G0 phase. During this time cells undergo changes that maximize their<br />
chances of survival. Predictably, this involves ashutting down of unnecessary metabolic<br />
pathways. Rates of protein synthesis are 300-fold lower in stationary phase yeast cells<br />
compared with those in the exponential phase (Fuge et al., 1994). Other changes occur<br />
that influence the metabolism <strong>and</strong> morphology of cells. Glycogen is dissimilated to<br />
provide carbon <strong>and</strong> energy for cellular maintenance. Some of the carbon is used to<br />
synthesize trehalose, which increases tolerance to stress (Section 12.5.7). Cell walls<br />
become thickened <strong>and</strong> more resistant to enzymatic degradation. Cells are more resistant<br />
to heat <strong>and</strong> desiccation. Cytoplasmic vacuolation becomes more extensive, presumably<br />
reflecting an increased requirement for turnover of intracellular protein.<br />
The metabolic changes associated with entry into the stationary phase are mediated at<br />
the gene level. Probably the RAS cyclicAMP signal transduction pathway controls entry<br />
into <strong>and</strong> exit from stationary phase. It is assumed that the cell has amethod for sensing<br />
the concentrations of nutrients <strong>and</strong> is able to initiate the necessary changes for entry into<br />
the stationary phase when it is apparent that the supply of an essential nutrient is about to<br />
disappearfrom themedium.Certainly,specificgenes appear tobeinducedjustbefore the<br />
stationaryphasecommences.Othersareexpressedonlyinstationaryphasecells.Therole<br />
of many of these remains to be elucidated (Werner-Washburne et al., 1996).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The stationary phase is transitory. The strategy of cells is to adapt to a physiological state<br />
that offers the longest survival time. During the early stationary phase, readily dissimilated<br />
sources of carbon such as glycogen are utilized. When these are depleted the cell will, if<br />
necessary, utilize structural macromolecules. If no nutrients become available the stationary<br />
phase terminates in death. The gross indication of this is a decrease in biomass<br />
concentration due to cell lysis; there is a loss of cellular integrity <strong>and</strong> a concomitant<br />
release of intracellular contents into the medium. The liberated cellular components can be<br />
utilized by other, still viable cells <strong>and</strong> even promote limited, cryptic growth.<br />
The different growth phases in a batch culture can be described mathematically. The<br />
rate of increase of biomass (x) with respect to time (t) can be expressed as:<br />
dx<br />
ˆ x 13:1<br />
dt<br />
Where is a constant termed the specific growth rate <strong>and</strong> has the unit of reciprocal time<br />
(h 1 ). During the lag <strong>and</strong> stationary phases it is equal to zero. It increases throughout the<br />
phase of accelerating growth <strong>and</strong> achieves a maximum value ( max) during the<br />
exponential phase.<br />
During exponential growth the rate of increase of biomass concentration becomes:<br />
dx<br />
dt ˆ maxx 13:2<br />
After integration <strong>and</strong> where x0 equals the initial biomass concentration, this yields the<br />
equation:<br />
lnx lnx0 ˆ maxt 13:3<br />
This can be rearranged to give the fundamental exponential growth equation.<br />
x ˆ x0e … maxt†<br />
From this equation it follows that a plot of lnx versus t is linear with a slope equal to max.<br />
The exponential phase of growth is conveniently quantified in terms of the time taken<br />
for the population to double in size. The doubling time (t D) is equal to:<br />
tD ˆ log102<br />
ˆ 0:693<br />
max<br />
max<br />
The number of doublings (n) which occur to reach a final biomass concentration of<br />
given by:<br />
is<br />
x<br />
n ˆ log2 13:6<br />
x0<br />
…x†<br />
n ˆ 3:32 log10<br />
x0<br />
At the end of the exponential phase, the growth rate becomes dependent upon the<br />
concentration of a growth-limiting substrate [S]. The relationship between growth rate<br />
<strong>and</strong> substrate concentration is expressed in the Monod equation:<br />
ˆ max<br />
‰SŠ<br />
Ks ‡ ‰SŠ<br />
The constant, Ks, is termed the saturation constant <strong>and</strong> is equal to the substrate<br />
concentration where equals half max. Plots of against substrate concentration show<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
13:4<br />
13:5<br />
13:7<br />
13:8
saturation kinetics. The inflexion point at which growth rate becomes independent of the<br />
substrate concentration varies with different strains.<br />
13.3.1 Brewery batch fermentations<br />
The pattern of growth in abrewery batch fermentation is similar to that shown in Fig.<br />
13.1. However, some aspects require further discussion. The initial yeast concentration in<br />
brewery fermentations is relatively high <strong>and</strong> the subsequent growth extent is modest.<br />
Thus, inoculation rates in atypical medium gravity (10ëP: SG c.1040) fermentation are<br />
approximately 3.0g/l wet weight of yeast, equivalent to around 0.6g/l cell dry weight, or<br />
in terms of cell numbers, approximately 1 10 7 cells/ml. During fermentation there are<br />
usuallynomorethantwotothreecelldoublings.Theconsequenceofthisisthattheyeast<br />
in the inoculum plays an important part in the subsequent fermentation.<br />
The inoculum is usually derived from aprevious fermentation <strong>and</strong> has been stored for<br />
aperiod before re-pitching. During the storage period the yeast is starving. It is usual to<br />
minimize the duration of the storage phase <strong>and</strong> to reduce metabolic activity by chilling to<br />
2±3ëC (35.6±37.4ëF). Nevertheless, the physiological state of the inoculum coupled with<br />
the conditions established at pitch influence subsequent patterns of growth. During the<br />
lagphasethepitchingyeast(whichisusuallysuspendedinbeer)adaptstoexposuretothe<br />
nutrients present in wort. In particular, the yeast is exposed to oxygen (see Sections 12.6<br />
<strong>and</strong> 12.7). The formation of sterol is accompanied by the concomitant mobilization of<br />
glycogen reserves (Fig. 13.2). There is alinear relationship between the quantities of<br />
glycogen dissimilated <strong>and</strong> sterol synthesized during fermentation (Quain <strong>and</strong> Tubb,<br />
1982). The formation of sterol, coupled with glycogen mobilization, is aprerequisite for<br />
the cells to move from the lag phase to the accelerating growth phase (Section 12.6).<br />
The events of very early fermentation in relation to the cell cycle are reviewed by<br />
Boulton <strong>and</strong> Quain (2001) (see also Section 11.7). Pitching yeast cells are in stationary<br />
phase (G 0) <strong>and</strong> unbudded. Transfer to wort induces a relatively synchronized shift to G 1<br />
phase <strong>and</strong> progress to START. In the first few hours there is no change in cell number or<br />
wort gravity. However, cell volume increases by approximately 20% <strong>and</strong> biomass falls by<br />
a similar percentage. During this time sterol synthesis takes place at the expense of cellular<br />
glycogen <strong>and</strong> molecular oxygen. Within six hours of pitching, almost 90% of the yeast<br />
population is budding, indicating that nearly all cells have passed into the S phase. The<br />
increase in budding index (relative number of budded cells) is very rapid, indicating a high<br />
degree of synchrony during very early fermentation. The achievement of high budding<br />
index corresponds with the onset of biomass increase <strong>and</strong> decrease in wort gravity.<br />
Glycogen<br />
(% cell dry wt.)<br />
40<br />
Glycogen<br />
0.12<br />
0.10<br />
0.08<br />
0.06<br />
Sterol<br />
0.04<br />
0.02<br />
0<br />
0 40 80 120 160<br />
0<br />
200<br />
Time (h)<br />
30<br />
20<br />
10<br />
Sterol<br />
(% cell dry wt.)<br />
Fig. 13.2 Changes in the intracellular concentrations of glycogen <strong>and</strong> total sterol in yeast<br />
measured during the course of a laboratory stirred fermentation using a lager yeast strain <strong>and</strong> 12 ëP<br />
all-malt lager wort (S. G. P. Durnin, unpublished results).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The increase in mean cell volume observed immediately after pitching is transitory. As<br />
the growth rate increases into the exponential phase, the mean cell volume decreases to a<br />
value similar to or slightly smaller than that seen at pitch. In the exponential phase, the<br />
population of yeast is made up of typically 25% multi-budded cells, 25% single-budded<br />
cells <strong>and</strong> 50% virgin daughters. The budding index also decreases from the initial high<br />
value <strong>and</strong> falls close to zero just after the mid-point of fermentation. Cell number<br />
increases to a constant value at the time that the budding index reaches a basal value.<br />
Biomass, measured as cell dry weight, declines slightly from the mid-point to the end of<br />
fermentation, reflecting the dissimilation of glycogen that occurs when growth has<br />
ceased.<br />
The decline in budding index to a value close to zero just after the mid-point of<br />
fermentation indicates that at this time growth is limited by the disappearance of a<br />
nutrient. In the majority of fermentations, made from all-malt wort, the limiting nutrient<br />
is probably oxygen <strong>and</strong>, by inference sterols <strong>and</strong>/or unsaturated fatty acids. In highly<br />
oxygenated worts, particularly those containing high levels of sugar adjuncts, nitrogen<br />
may be the limiting substrate.<br />
13.3.2 Effects of process variables on fermentation performance<br />
The batch growth curve is influenced by a number of physical parameters. These are;<br />
temperature, pitching rate, dissolved oxygen concentration, wort concentration <strong>and</strong><br />
pressure. Adequate control of these parameters is essential to ensure consistent<br />
fermentation performance <strong>and</strong> beer quality. Specific effects can be ascribed to<br />
modulating each parameter in isolation, nevertheless, they are all to some extent<br />
interdependent. For example, increasing fermentation temperature reduces oxygen<br />
solubility. Similarly, a decrease in wort concentration increases oxygen solubility. When<br />
examining the effects of altering a single parameter it is essential to ensure that all others<br />
remain constant. Some of the effects described will be observed only if the parameters are<br />
varied over very wide ranges. These extremes will not be seen in real brewery<br />
fermentations because they are beyond the limits that would be used in <strong>practice</strong>, but they<br />
serve as useful illustrations of the underlying principles.<br />
An increase in the fermentation temperature reduces the time taken to attenuate the<br />
wort (Fig. 13.3). Thus, temperature exerts its effects primarily on yeast metabolic rate. It<br />
PG<br />
50<br />
40<br />
30<br />
20<br />
10<br />
25°C<br />
20°C<br />
15°C<br />
10°C<br />
0<br />
0 40 80 120<br />
Time (h)<br />
160 200 240<br />
Fig. 13.3 Effect of temperature on fermentation rate. Fermentations were performed at the<br />
temperatures indicated using pilot scale 8 hl cylindroconical fermenters, a lager yeast strain, pitched<br />
at a rate of 12 10 6 viable cells/ml into an all-malt lager wort with an OG of 1050 (12.5 ëP) <strong>and</strong> an<br />
initial dissolved oxygen concentration of 25 mg/l (Boulton <strong>and</strong> Box, unpublished data).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
has little effect on the extent of yeast growth during fermentation. Most brewing strains<br />
have an optimum temperature for growth between 30 <strong>and</strong> 34ëC (86±93.2ëF) but<br />
fermentations are maintained at lower temperatures. Superficially, the use of elevated<br />
temperatures for minimizing vessel turn-round times appears to be an attractive strategy.<br />
There are several disadvantages. At high temperatures, especially in cylindroconical<br />
vessels, fermentation is very vigorous. Rates of CO2 evolution are very rapid <strong>and</strong> losses<br />
of volatiles, via gas stripping, may become unacceptable. Indeed, beer losses due to<br />
uncontrollable fobbing may occur. To avoid this it may be necessary to operate<br />
fermenterswithalargefreeboard.Theconsequentreductioninvesselproductivecapacity<br />
offsets some of the gains made by shorter vessel residence times. At the end of<br />
fermentation, the time taken to chill is directly proportional to the fermentation<br />
temperature.<br />
Generally, ale fermentations are performed at higher temperatures (18 22ëC;<br />
64.4 71.6ëF) compared with lager types (8 15ëC; 48.4 59ëF). These temperatures<br />
are chosen largely because they produce adesired spectrum of yeast-derived beer flavour<br />
components. Deviating from the normal temperature ranges may produce totally<br />
unacceptable shifts in beer flavour. Increasing the fermentation temperature leads to<br />
increases in the concentrations of higher alcohols <strong>and</strong> esters during fermentation (see<br />
Sections 12.10.3; 12.10.4). These flavour changes are particularly undesirable in the case<br />
of pale lager beers.<br />
The concentration of oxygen supplied in wort at the start of fermentation is one of the<br />
primary regulators of yeast growth (see Sections 12.6, 12.7). Although wort composition<br />
is also influential, there is adirect correlation between the initial dissolved oxygen<br />
concentration<strong>and</strong>theextentofyeastgrowth.Elevation inoxygenconcentration resultsin<br />
an increased primary fermentation rate. Caution must be exercised in using very high<br />
oxygen concentrations as a means of reducing fermentation times. High oxygen<br />
concentrations produce rapid primary fermentations but this may be at the expense of<br />
ethanol yield. Theincreasedavailabilityofoxygen promoteshigh rates ofyeastgrowth at<br />
the expense of sugar, which would otherwise be available for ethanol formation (Fig.<br />
13.4). Since oxygen has adirect effect on the extent of yeast growth, it would be<br />
predicted that it would also influence the concentrations of flavour metabolites produced<br />
as aresult of yeast growth. This is indeed the case. With many yeast/wort combinations,<br />
but not all, there is an inverse correlation between initial oxygen concentration <strong>and</strong> beer<br />
ester levels.<br />
At low values there is adirect correlation between the yeast pitching rate <strong>and</strong> the rate<br />
of primary fermentation <strong>and</strong> the extent of yeast growth. At higher pitching rates an<br />
inflection point is reached beyond which yeast growth decreases <strong>and</strong> there is no further<br />
increase in primary fermentation rate (Fig. 13.5). These observations reflect the<br />
interrelationships between yeast pitching rate <strong>and</strong> wort composition, in particular the<br />
initial dissolved oxygen concentration. Brewery fermentations differ from most other<br />
commercial fermentations in that the size of the inoculum is high relative to the extent of<br />
subsequent growth, so the ratio between the initial yeast concentration <strong>and</strong> available<br />
nutrients becomes asignificant factor.<br />
At low pitching rates (below the inflection point), there is adirect relation with<br />
attenuation rate. Total yeast growth is probably limited by oxygen-dependent lipid<br />
synthesis. Thus, there is asurplus of oxygen relative to the number of yeast cells present.<br />
However oxygen is also expended in other non-lipogenic pathways (Section 12.5). The<br />
small initial yeast population becomes lipid replete but subsequent daughter cells are<br />
limited by the onset of anaerobiosis. Below the inflection point, yeast growth extent <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Ethanol. conc. (g/l)<br />
Fermentation rate<br />
(time to half-gravity, h)<br />
80<br />
Yeast conc. 5<br />
60<br />
4<br />
40<br />
Ethanol 3<br />
2<br />
20<br />
1<br />
0<br />
0 10 20 30<br />
0<br />
40<br />
(Dissolved oxygen) (mg/l)<br />
100<br />
80<br />
Deg.<br />
fermented<br />
60<br />
60<br />
Fermentation<br />
40<br />
40<br />
20<br />
rate<br />
20<br />
0<br />
0 10 20 30<br />
0<br />
40<br />
Fig. 13.4<br />
(Dissolved oxygen) (mg/l)<br />
Effect of varying initial dissolved oxygen concentration on fermentation rate (expressed<br />
as time to reach half-gravity) <strong>and</strong> the formation of ethanol <strong>and</strong> new yeast growth during the<br />
fermentation of a 15 ëP lager wort with a lager yeast strain. The temperature was 11 ëC <strong>and</strong> the<br />
pitching rate was 15 10 6 viable cells/ml (redrawn from Bamforth et al., 1988).<br />
Fermentation rate<br />
(time to half-gravity, h)<br />
60<br />
40<br />
20<br />
Fermentation<br />
rate<br />
Growth<br />
0<br />
0<br />
0 10 20 30 40<br />
Pitching rate (viable cells/ml × 10 6 )<br />
Fig. 13.5 Effect of pitching rate on fermentation rate (time to half-gravity) <strong>and</strong> new yeast growth<br />
(weight of crop ± weight of yeast pitched) for 8 hl pilot scale all-malt 15 ëP lager fermentations. The<br />
initial wort oxygen concentration was 25 mg/l <strong>and</strong> the pitching rate was 12 10 6 viable cells/ml<br />
(re-drawn from Boulton <strong>and</strong> Quain, 2001).<br />
pitching rate are directly related. With increasing pitching rate a point is reached where<br />
growth is limited by the nutrient supply available to each yeast cell. The quantity of<br />
oxygen supplied per yeast cell is probably the limiting nutrient. Attenuation rates remain<br />
high because of the high yeast population. The patterns of growth <strong>and</strong> fermentation rates<br />
shown in Fig. 13.5 will be modified by other parameters such as the initial wort oxygen<br />
concentration <strong>and</strong> availability of other nutrients. Thus, at very high initial oxygen<br />
concentrations the inflection point at which growth begins to decline will be shifted<br />
towards the right. The inflection will still occur because another nutrient will eventually<br />
become growth limiting.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
30<br />
20<br />
10<br />
Yeast growth (g/l wet wt.)<br />
Yeast conc. (g/l dry wt.)<br />
Degrees fermented
SG<br />
1050<br />
1040<br />
1030<br />
1020<br />
1010<br />
1000<br />
50% increase<br />
30% increase<br />
20% increase<br />
0 2 4 6 8<br />
Fermentation time (days)<br />
Fig. 13.6 The relationship between fermentation time <strong>and</strong> increasing the original gravity of wort<br />
from 1033.3 (8.3 ëP) in cylindroconical fermenters (redrawn from Hough et al., 1982).<br />
Increasing the wort concentration, without changing any other parameters, results in<br />
an increase in fermentation time. However, providing the pitching rate <strong>and</strong> wort dissolved<br />
oxygen concentration are increased pro rata, there is only a small increase in<br />
fermentation times (Fig. 13.6). This forms the basis of high-gravity brewing in which<br />
concentrated worts are fermented <strong>and</strong> subsequently diluted to sales gravity. Fermentation<br />
profiles are influenced by wort composition, such as variations in the spectrum of<br />
fermentable sugars. Some of these effects are strain-specific. Lager strains utilize<br />
maltotriose more rapidly than ale strains (Stewart et al., 1995). It would be predicted,<br />
therefore that a change in the concentration of this sugar would have different effects<br />
depending on the nature of the yeast strain. At limiting concentrations, fermentation rates<br />
are proportional to the concentration of -amino nitrogen (Fig. 13.7). In an all-malt wort<br />
the supply of -amino nitrogen should be adequate <strong>and</strong> this dependence should not be<br />
observed. Deficiencies may arise in high-gravity worts made with a high proportion of<br />
non-malt adjunct.<br />
Several distinct effects can be ascribed to pressure. Yeast is subjected to an osmotic<br />
pressure the magnitude of which is dependent on the concentration of the wort. In highgravity<br />
worts the osmotic pressure may be as high as 4 10 6 Pa (Owades, 1981). Yeast<br />
Fermentation rate (h)<br />
120<br />
100<br />
80<br />
60<br />
40<br />
20<br />
0<br />
0 20 40 60 80<br />
α-Amino nitrogen (mg/100 ml)<br />
Fig. 13.7 Relationship between fermentation time <strong>and</strong> -amino nitrogen of the wort (redrawn<br />
from Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
cells appear unaffected by these pressures providing they are not subjected to abrupt<br />
changes. Osmotic pressures up to 10 8 Paare tolerated by yeast (Gervais et al., 1992). The<br />
osmotic potential of aqueous media is directly proportional to the concentration of<br />
dissolved solutes. This parameter is inversely related to the concentration of available<br />
water <strong>and</strong> is referred to as the water activity (A w). All organisms are able tolerate agiven<br />
range of Aw. This parameter appears to be of greater significance to brewing yeast than<br />
osmotic potential per se (Chapter 12; Section 12.3.1).<br />
In fermenters yeast is subjected to a hydrostatic pressure that is a function of the height<br />
of the vessel. In addition, if the exit of CO2 produced during fermentation is restricted,<br />
the vessel will become pressurized. Very high pressures are deleterious to yeast. Very<br />
high-pressure treatments (> 10 8 Pa; 100 Bar) have been used to sterilize some foods<br />
where heat treatments cause deleterious flavour changes. In similar fashion to the dual<br />
effects of osmotic pressure <strong>and</strong> water activity, pressurization of fermenters leads to a<br />
concomitant increase in the concentration of dissolved CO2. The latter may have a greater<br />
effect on yeast than pressure alone (Thibault et al., 1987).<br />
Moderate top-pressurization of fermenters (1.2 2.0 10 5 Pa, 1.2 2.0 Bar) has been<br />
used to reduce yeast growth <strong>and</strong> control fobbing. This approach has been used where<br />
changing other parameters, such as the use of very concentrated worts or elevated<br />
temperature, has resulted in unacceptable perturbations in the formation of flavour<br />
metabolites due to increased yeast growth (Nielsen et al., 1986, 1987). The effects due<br />
to pressure are strain-specific. Using a spheroconical fermenter Posada (1978) reduced<br />
yeast growth by the application of pressure. However, depending on the yeast strain this<br />
was accompanied by both upward <strong>and</strong> downward shifts in the concentrations of various<br />
flavour metabolites. Miedaner (1978) described a protocol for high-temperature<br />
fermentations in which vessels were allowed to pressurize to approximately 1.8<br />
10 5 Pa (1.8 Bar) at a point when the wort was approximately 50% attenuated. This<br />
resulted in a reduction in yeast growth <strong>and</strong> levels of higher alcohols compared to<br />
unpressurized lower-temperature fermentations. Comparative analysis of beers suggested<br />
that this was at the expense of some damage to yeast. Thus, the pH of the trial<br />
beer was higher <strong>and</strong> head retention values were reduced, possibly due to the presence of<br />
greater than normal concentrations of short chain fatty acids. Both of these effects<br />
suggest yeast autolysis.<br />
13.4 Yeast ageing<br />
Most analyses of yeast growth consider the dynamics of whole populations of cells.<br />
Biomass is estimated by cell number or cell weight. Where population analyses are<br />
performed they are often rudimentary, for example assessments of the proportions of<br />
living <strong>and</strong> dead cells. Where the viable population is analysed, as in the so-called vitality<br />
tests, most assessments are based on mean values of all the cells within the sample. These<br />
approaches assume that there is no heterogeneity within yeast populations. Budding<br />
yeasts differ from organisms that reproduce by binary fission in that individual cells have<br />
a finite life span determined by the number of times that it buds <strong>and</strong> its DNA is replicated.<br />
Replicative age is distinct from chronological age. When cells are unable to bud further<br />
they become senescent <strong>and</strong> ultimately they die (Jazwinski, 1999; Powell et al., 1999).<br />
Replicative age of yeast is measured relatively easily in that each budding event leaves a<br />
characteristic scar on the cell wall. This can be visualized with fluorophores such as the<br />
dye, calcofluor. The maximum number of times individual strains are able to bud is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
strain-specific <strong>and</strong> varies between about 15 <strong>and</strong> 40, usually +/ about 10. This number is<br />
termed the Hayflick limit (Hayflick, 1965).<br />
Ageing is accompanied by morphological changes. The cells gradually take on a<br />
wrinkled <strong>and</strong> granular appearance. There is apositive correlation between cell age <strong>and</strong><br />
cell volume (Barker <strong>and</strong> Smart, 1996), <strong>and</strong> for alager strain aplot of cellular size versus<br />
age was linear. There was asixfold difference between the size of young mother <strong>and</strong><br />
senescent cells. Typically, generation times increase sharply just before the onset of<br />
senescence. Cells entering the senescent phase commonly fail to separate from daughter<br />
cells. The senescent phase culminates in death. In higher eukaryotes this process is<br />
termed apoptosis or programmed cell death. It is not ar<strong>and</strong>om process. In multicellular<br />
eukaryotes cells are continually dying <strong>and</strong> being replaced. Death appears to be under<br />
genetic control <strong>and</strong> occurs when cells have degenerated, possibly as aconsequence of<br />
damage by reactive oxygen radicals (Madeo et al., 1999). The apoptotic pathway has not<br />
been positively identified in yeast but it is possible that cells undergo suicide for similar<br />
reasons. In yeast, death is accompanied by autolysis, literally self-digestion. The process<br />
involves loss of membrane integrity <strong>and</strong> the breakdown of cellular macromolecular<br />
components by avariety of hydrolytic enzymes.<br />
Yeast cell ageing may serve as amodel for the same process in higher eukaryotes,<br />
including man. In brewing it may also have relevance to fermentation management.<br />
<strong>Brewing</strong> yeastpopulations may beheterogeneouswith respect toage.Given the relation<br />
between replicative age <strong>and</strong> cell size it might be supposed that old <strong>and</strong> young cells<br />
would have different sedimentation characteristics in the cones of cylindroconical<br />
fermenters. This was the case in a2000hl fermenter (Deans et al., 1997). Yeast cells at<br />
the bottom of the cone were, on average, older than those at the top. The fermentation<br />
performance of the older yeast fraction was significantly poorer compared to younger<br />
cells. This prompted the suggestion that the first portion of crops from these vessels<br />
should be discarded. In another investigation (Quain et al., 2001) contrary results were<br />
obtained. In this case yeast cells of different average replicative age were fractionated<br />
using sucrose gradient centrifugation. The fermentation performance of each fraction<br />
was compared. The larger older cells produced fermentations with faster attenuation<br />
rates compared to the smaller younger cells. It is difficult to reconcile these<br />
diametrically opposite results. In both investigations there were significant differences<br />
between fermentation performance <strong>and</strong> replicative age suggesting that more detailed<br />
investigations are needed.<br />
13.5 Yeast propagation<br />
Theoretically there is no limit to the number of times that yeast may be serially cropped<br />
<strong>and</strong> re-pitched. Some traditional breweries, particularly those using top-cropped ale<br />
strains, have followed this <strong>practice</strong> for many years without interruption. Most modern<br />
breweries periodically introduce new cultures of yeast of guaranteed identity <strong>and</strong> purity<br />
derived from laboratory stocks. This is done for several reasons. Where several yeast<br />
strains are used within the same brewery low levels of contamination are inevitable <strong>and</strong><br />
there is aconstant threat of contamination with wild yeasts <strong>and</strong> spoilage bacteria. Acid<br />
washing(Section17.6)shouldcontrolspoilagebacteria,butithasnoeffectonwildyeast.<br />
With prolonged serial fermentation the characteristics of the production yeast may<br />
change due to genetic instability. Petite mutants, which lack the ability to form functional<br />
mitochondria, are very common in brewing yeasts. Petites ferment abnormally (Ern<strong>and</strong>es<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
et al., 1993). Other types of genetic instability have been observed in brewing yeast,<br />
particularly changes from arelatively non-flocculent to aflocculent character (Section<br />
11.8.2).Bottom-croppingyeastfromcylindroconical fermenterscanselectforthesemore<br />
flocculent variants. This effect may be exacerbated with repeated serial fermentation.<br />
Bottom cropping has been associated with other undesirable effects. Prolonged serial<br />
cropping can result in a progressive enrichment of pitching slurries with trub <strong>and</strong> other<br />
non-yeast particulate matter. Not only is this trub added to the next fermentation, it<br />
results in an underestimate in the calculation of pitching rate where this is determined by<br />
measurement of spun solids. Bottom cropping may tend to select for larger <strong>and</strong> therefore<br />
older cells (Smart <strong>and</strong> Whisker, 1996; Deans et al., 1997. This effect may contribute to a<br />
gradual decline in the performance of brewing yeast with generational age.<br />
Typically, yeast is serially repitched between 5 <strong>and</strong> 20 times before disposal. This<br />
wide range reflects the importance that individual brewers place on the need to introduce<br />
new cultures. Conversely, it reflects the threat that individual brewers consider is posed<br />
by prolonged serial re-pitching. Propagation attracts both capital <strong>and</strong> revenue costs. It is<br />
commonly asserted that newly propagated yeast does not produce st<strong>and</strong>ard fermentation<br />
performance or beer, so there is a natural reluctance to propagate frequently, especially if<br />
existing yeast lines are performing in a satisfactory manner. The suggestion that newly<br />
propagated yeast performs poorly is unproven <strong>and</strong> may simply reflect less than ideal<br />
propagation plant. The decision to introduce a new culture should be based upon<br />
microbiological <strong>and</strong> performance testing of existing yeast. The process should be<br />
managed so that a new culture is introduced when experience suggests that older cultures<br />
will be approaching the end of their useful lifecycles.<br />
Yeast propagation has three elements. Firstly, there is a need to maintain stock<br />
cultures. Secondly, the stock culture must be used to generate a laboratory culture of a<br />
scale sufficient to pitch the first brewery culture. Thirdly, the yeast must be propagated<br />
within the brewery to grow an amount sufficient to pitch the first production scale<br />
fermentation.<br />
13.5.1 Maintenance <strong>and</strong> supply of yeast cultures<br />
The complexity of the system used for the maintenance of stock cultures depends on the<br />
size of the operation. It must be a quality assured system in which cultures of guaranteed<br />
identity <strong>and</strong> purity are delivered to the brewery. Several levels of complexity are possible.<br />
Small breweries using a single yeast strain may hold stock cultures at independent thirdparty<br />
institutions such as the various national collections of yeast cultures. The onus for<br />
guaranteeing the quality of the supplied yeast is placed, at a cost, on the institution. This<br />
approach has the advantage of simplicity. Many brewers maintain their own strains. This<br />
can range from a requirement to look after a single strain at one brewery to multiple<br />
strains supplied to several breweries. Where a single company has to supply several<br />
satellite breweries <strong>and</strong> possibly a number of franchise breweries with a number of yeast<br />
strains it is convenient to have a dedicated central facility. This facility replaces the thirdparty<br />
operators <strong>and</strong> takes on the task of quality assurance of cultures <strong>and</strong> their supply.<br />
The satellite breweries have the much reduced burden although still essential task of<br />
assuring the supply of cultures from their own brewery laboratories into propagator <strong>and</strong><br />
thence production. Alternatively, the central facility may undertake propagation <strong>and</strong><br />
supply bulk yeast to breweries.<br />
There is a need to store cultures for long periods in such a way that they remain pure,<br />
at high viability <strong>and</strong> not subject to genetic change. Several methods are used <strong>and</strong> they<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
fulfil these criteria with varying degrees of success. The simplest method is by periodic<br />
sub-culture using agar slopes (slants). These consist of small bottles, typically<br />
containing around 10ml of asuitable nutrient medium, solidified with agar. In order<br />
to maximize the surface area, the agar is allowed to solidify with the bottle placed at a<br />
slant, hence the name. The agar is inoculated with apure culture of yeast <strong>and</strong> incubated<br />
to provide aprofuse layer of growth on the surface of the agar. Slope cultures are stored<br />
at 2 4ëC (35.6 39.2ëF) to minimize yeast metabolic activity <strong>and</strong> prolong the<br />
maximum storage period. Periodically slopes are sub-cultured by transfer to fresh<br />
medium. This approach is simple <strong>and</strong> inexpensive <strong>and</strong>, providing skilled personnel<br />
perform it, should not result in loss of purity. It has the major disadvantage that while<br />
metabolism is slowed by cold storage it is not stopped. Prolonged storage results in loss<br />
of viability, nevertheless cultures can be held in this way for 4 6months. To allow a<br />
margin for safety, slopes should be sub-cultured every three months. The most serious<br />
disadvantage of this method is that over long periods of time <strong>and</strong> following multiple<br />
sub-culturing genetic drift <strong>and</strong> selection of non-st<strong>and</strong>ard variants has occurred (Kirsop,<br />
1991).<br />
More sophisticated storage methods seek to slow down metabolism further than can<br />
be achieved by chilling alone <strong>and</strong> thereby prolong storage times. Apopular method is<br />
that of lyophilization or freeze-drying. Cultures are rapidly frozen followed by drying<br />
under vacuum such that water is removed by sublimation. The process is performed in<br />
glass ampoules, which are sealed when drying is complete. These cultures can be safely<br />
stored for several months. Reactivation is achieved by breaking the ampoule <strong>and</strong><br />
transferring the dried biomass to fresh liquid medium. This method is widely used but it<br />
has serious shortcomings. Freeze drying results in alarge overall reduction in viability.<br />
The fraction that remains viable appears to do so for several months, thereafter but<br />
usually up to 95% of the original cells die during drying. More worryingly, the viable<br />
fraction may undergo some degree of genetic disruption during freeze drying (Russell<br />
<strong>and</strong> Stewart, 1981).<br />
The death <strong>and</strong> deterioration that accompanies freeze-drying is probably caused by the<br />
formation of intracellular ice crystals (Morris et al., 1988). In other industries, where the<br />
use of dried yeast is commonplace, yeast is cultivated in amanner that manipulates<br />
physiology to render the cells less susceptible to the rigours of drying. Thus, cells are<br />
encouraged to accumulate trehalose, a well-recognized stabilizer of biological<br />
membranes (Section 12.5.7) <strong>and</strong> protectants are added during processing. Guldfeldt<br />
<strong>and</strong> Piper (1999) reported that ale <strong>and</strong> lager strains stored in liquid nitrogen retained<br />
100% viability after seven months storage. The same strains dried under vacuum in the<br />
presence of glucose suffered losses in viability between 10 <strong>and</strong> 99.6%. When yeast was<br />
subjected to an osmotic shock, in the form of exposure to sorbitol (20%w/v) prior to<br />
drying the viability decrease was reportedly much less.<br />
The adoption of dried yeast as the method of choice for bakers has also been<br />
recommended to brewers (Debourg <strong>and</strong> Van Nedervelde, 1999). Fermentation<br />
performance <strong>and</strong> beer quality at pilot <strong>and</strong> production scale were comparable with those<br />
produced using conventional pitching yeast. Dried yeast is used in parts of Africa for<br />
making African-style opaque beer (Chapter 16). It is suggested that the use of dried yeast<br />
mightreplaceconventionalbrewerypropagation.Itsuseobviatestheneedforpropagation<br />
<strong>and</strong>thesubsequentstorageofcroppedpitchingyeast.Theoretically,driedyeastshouldbe<br />
ofaconsistentphysiologicalcondition<strong>and</strong>notsufferthevariabilitythatmightbethecase<br />
with conventional pitching yeast. On the debit side, bakers' yeast destined for drying is<br />
grown so that all cells are fully respiratory (Section 13.5). Such cells should be sterol<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
eplete <strong>and</strong> have little or no requirement for oxygenation of worts. Proof is still required<br />
thatbrewingyeastinthisstatewillgivest<strong>and</strong>ardfermentationperformance<strong>and</strong>beerwith<br />
anormalvolatilespectrum.Nevertheless,theuseofdriedyeastisanattractiveoption<strong>and</strong><br />
may well find application where an infrequent supply isneeded such as might be the case<br />
with franchise brewing, pub brewing <strong>and</strong> home brewing.<br />
The most effective (<strong>and</strong> most expensive) method of preservation <strong>and</strong> storage of yeast<br />
cultures is freezing in liquid nitrogen ( 196ëC; 320.8ëF). Cultures must be frozen in a<br />
controlled manner but once this is achieved the storage potential is measured in years.<br />
Furthermore,nochangesingenotypehavebeenreported(Kirsop,1991;Quain,1995).As<br />
with the method for bulk drying bakers' yeast, the cells to be frozen are grown under<br />
oxidative conditions. Prior to freezing yeast is suspended (c. 30% wet wt/vol) in a<br />
medium containing glycerol (5% v/v) as acryoprotectant. To maintain high viability the<br />
rate at which the temperature is lowered must be regulated. The temperature must be<br />
reduced slowly during the first phase, typically two hours from 20ëC to 30ëC (68 to<br />
22ëF). After this the yeast is rapidly cooled to 196ëC ( 320.8ëF) by immersion in<br />
liquid nitrogen. The underlying principle of the process is that during the initial slow<br />
phase the suspending medium freezes first. Consequently, the osmotic potential of the<br />
suspending medium increases, a phenomenon assisted by the presence of the<br />
cryoprotectant. This causes intracellular water to be released into the medium <strong>and</strong> in<br />
consequence the cells shrink. The gradual shrinkage <strong>and</strong> dehydration prevents<br />
intracellular ice crystal formation.<br />
Yeast is conveniently frozen in colour-coded sealed straws <strong>and</strong> held in purpose-built<br />
liquid nitrogen refrigerators. Yeast is recovered by plunging the straws into warm (37ëC)<br />
water, broaching using sterile scissors then transferring the thawed slurry to sterile liquid<br />
medium. This is used to prepare amaster culture from which slope cultures are prepared.<br />
At the same time checks of strain purity <strong>and</strong> possibly identity are made. The slopes are<br />
distributed to breweries for use in propagation. An ISO 9000 quality assured system for<br />
yeast preservation using liquid nitrogen, recovery of cultures <strong>and</strong> distribution to satellite<br />
breweries is described by Quain (1995).<br />
13.5.2 Laboratory yeast propagation<br />
The aim of laboratory propagation is to grow apure culture of yeast of sufficient volume<br />
to pitch the first brewery scale propagation vessel. The process is performed using<br />
traditional microbiological techniques <strong>and</strong> mainly glass apparatus, preferably by skilled<br />
personnel. The importance of this stage is often underestimated. Ensuring that the culture<br />
ispureisofparamountimportance.Althoughitiscustomarytoconfirm thatthecultureis<br />
free from contamination the test results may not be obtained until after the first brewery<br />
stage has been pitched.<br />
To generate the terminal laboratory propagation cultureit isnecessary togrow aseries<br />
of intermediate cultures of progressively increasing volume. Atypical protocol is shown<br />
in Fig. 13.8. Scale up factors are usually around 1:10. The initial stages use ageneralpurpose<br />
yeast medium such as yeast extract, peptone glucose (YEPG). The terminal<br />
phase uses sterile brewery wort. The whole process takes around two weeks. Growth of<br />
yeast is promoted by continuous aeration. Terminal yeast counts should be within the<br />
range 150±200 10 6 cells per ml at a viablity greater than 98%. The sub-terminal culture<br />
is carried out using an aspirator flask. The culture is aerated via a glass sinter fitted with a<br />
sterile gas filter. Good rates of oxygen transfer are encouraged by mechanical agitation<br />
using a magnetic follower. The outlet on the aspirator is fitted with silicone tubing <strong>and</strong> a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Slope culture<br />
10 ml YEPG<br />
100 ml YEPG<br />
2l YEPG<br />
20l sterile wort<br />
Brewery seed vessel<br />
Wash all growth off slope<br />
with sterile YEPG medium<br />
Static incubation<br />
for 24h at 20°C<br />
Incubate on flask shaker<br />
for three days at 25°C<br />
Incubate for three days<br />
at 25°C with continuous<br />
agitation <strong>and</strong> aeration<br />
Incubate for three days<br />
at 25°C with continuous<br />
agitation <strong>and</strong> aeration<br />
Fig. 13.8 Protocol for the laboratory phase of yeast propagation. All stages except the terminal<br />
step use a semi-defined medium (YEPG: yeast extract, 5 g/l; peptone, 10 g/l; glucose, 20 g/l).<br />
coupling for attachment to the inoculation port on the apparatus used for growing the<br />
terminal laboratory culture. The latter piece of apparatus must fulfil three functions. It<br />
must be able to withst<strong>and</strong> autoclaving during sterilization of the wort. It must be suitable<br />
for growing pure cultures of yeast to high concentration without risk of contamination. It<br />
must be suitable for transferring yeast from the laboratory to the brewery, therefore it<br />
must be of robust construction to withst<strong>and</strong> the rigours of the production environment.<br />
Suitable apparatus for carrying out the terminal phase of laboratory propagation is<br />
shown in Fig. 13.9. The flask has atotal capacity of approximately 25 litres <strong>and</strong> is<br />
constructed from stainless steel. Several portstraverseatop plate thatcan beremoved for<br />
filling with wort <strong>and</strong> for cleaning. These ports are attached to lines for inoculation,<br />
sampling, gas inlet <strong>and</strong> outlet <strong>and</strong> transfer to the brewery seed vessel. During the growth<br />
phase the wort is aerated using air delivered via asterile gas filter <strong>and</strong> astainless steel<br />
sinter. The wort is agitated using amagnetic follower. Transfer of the culture to the<br />
brewery seed vessel is achieved using agas cylinder to provide motor gas delivered via<br />
the exhaust line. All gas lines are protected with sterile filters <strong>and</strong> all other lines are fitted<br />
with connectors, which are wrapped during sterilization. These are used to make aseptic<br />
connections when transfers are made.<br />
13.5.3 Brewery propagation<br />
The culture supplied by the laboratory is pitched into asmall seed tank <strong>and</strong> thence<br />
through afurther series of vessels of increasing volume until sufficient yeast has been<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Sterile wrapped<br />
connection to<br />
brewery seed vessel<br />
Top plate clamped to<br />
flask body via sterilizable<br />
silicone rubber O-ring<br />
Stainless<br />
steel flask<br />
Sterile<br />
gas filters<br />
Sterile wrapped<br />
connection for<br />
addition of inoculum<br />
Exhaust gas outlet/<br />
transporter gas inlet<br />
Aeration inlet attached<br />
to stainless steel c<strong>and</strong>le<br />
Ports in head plate fitted<br />
with sterile seals<br />
Magnetic follower<br />
Magnetic stirrer<br />
Fig. 13.9 Components of apparatus suitable for carrying out the terminal growth phase of<br />
laboratory yeast propagation. The flask has an operating volume of 20 l <strong>and</strong> a total volume of 25 l.<br />
generated to pitch the first production fermentation. From amicrobiological st<strong>and</strong>point<br />
propagation is fraught with risk. Any contamination at this stage will have profound<br />
adverse consequences. The design of the plant, its operation <strong>and</strong> its internal finish must<br />
be to the highest hygienic st<strong>and</strong>ards. The propagation plant should be located within a<br />
room designed to minimize the risk of contamination. Access is restricted to essential<br />
personnel, all internal surfaces are made from hygienic materials, the air in the room is<br />
filtered <strong>and</strong> the atmosphere maintained under aslight positive pressure.<br />
Wort is used as the propagation medium. Preferably, it should be of the same quality<br />
as that used in fermentation. This is usually not essential from the point of yeast growth<br />
but as the spent wort will eventually be pitched with the yeast it should match the wort it<br />
is pitched into. Yeast should not be propagated using high-gravity worts (Cahill <strong>and</strong><br />
Murray, 2000). For ale <strong>and</strong> lager strains propagated on 7.5, 11.5 <strong>and</strong> 17.5ëP worts there<br />
wasaprogressiveincreaseinthemeancellsize<strong>and</strong>decreaseinviability.Bearinginmind<br />
that ethanol is toxic <strong>and</strong> that actively growing cells are the most susceptible this is<br />
unsurprising (Section 12.5.9). Prior to inoculation the wort <strong>and</strong> vessel must be sterilized.<br />
The vessel may be sterilized empty <strong>and</strong> then filled with wort from the hot side of the<br />
paraflow. Preferably, propagation vessels are provided with a means of boiling wort in<br />
situ via steam jackets or direct injection of steam. Yeast growth is encouraged by<br />
provision of air or oxygen. Gas inlet <strong>and</strong> exhaust lines must be fitted with microbiological<br />
filters. Sampling <strong>and</strong> inoculation ports must be capable of aseptic operation, preferably<br />
sterilized by steaming.<br />
The performance of the first fermentation should fall within the normal range in terms<br />
of duration <strong>and</strong> the extent of yeast growth. The resultant beer should be within<br />
specification. These goals are not always achieved. Traditional propagators operate on<br />
the principle that the yeast should, as far as possible, be in the same physiological<br />
condition as pitching yeast. The assumption is that this will ensure that the newly<br />
propagated yeast will produce st<strong>and</strong>ard fermentation performance when it is transferred<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
into the first wort. This is achieved by limiting the quantity of oxygen supplied to yeast<br />
during the growth phase <strong>and</strong> controlling the temperature at the same value or slightly<br />
higher than that used for fermentation. Traditional propagators tend to be of similar<br />
construction to fermenters. Thus, mechanical agitation is not provided <strong>and</strong> during the<br />
growth phase, aeration is not continuous. The downside to this approach is that the yield<br />
of yeast is low, usually no more than would be obtained from aconventional wort<br />
fermentation (typically, 50±60 10 6 cells per ml). The low yields associated with<br />
traditional propagation systems usually require several steps to generate sufficient<br />
biomass to pitch the first fermentation. Furthermore, scale-up factors between each phase<br />
are modest, not more than 1:5. Atypical regime would require four stages of 1.5, 6.0,<br />
30.0 <strong>and</strong> 150hl, respectively to generate sufficient yeast to pitch up to 800hl (500 UK<br />
barrels) of wort. The duration of the entire process, including the laboratory phase, takes<br />
several weeks. Thus, the process is slow, the use of multiple vessels is costly to install<br />
<strong>and</strong> operate. The risks of contamination are proportional to the number of vessels used.<br />
In modern breweries, which may use several yeast strains, traditional propagation<br />
systemscannoteasilysatisfytherequirementsforyeast.Anadditionalcomplicationisthe<br />
use of very large fermenters. Commonly, very large fermenters have been installed<br />
without concomitant upgrading of the propagation plant. To overcome this problem it<br />
may be necessary to continue using smaller fermenters in parallel with the larger ones<br />
simply to generate sufficient yeast. Alternatively, large fermenters can be part-filled<br />
when pitched with newly propagated yeast.<br />
To satisfy the increased need for propagation in large breweries it has been necessary<br />
to introduce methods for increasing the yield of yeast <strong>and</strong> accelerating the process.<br />
Consider abrewery producing beers with an average fermenter turn-round time of 12<br />
days, using six yeast strains. Assuming yeast is serially pitched for 15 generations after<br />
which time anew culture is introduced, the total life-time of each yeast culture would be<br />
(12 15) ˆ180 days. Assuming apropagation regime which required 35 days to<br />
complete (14 days laboratory stage +21 days brewery stage) the total time required to<br />
accommodate all six yeast strains would be 210 days. Clearly, in this case it would be<br />
necessary to duplicate some of the propagation plant in order to meet the needs of the<br />
brewery.<br />
Growing the yeast at a temperature higher than that used during fermentation<br />
accelerates the propagation process. High yields can be obtained by ensuring that<br />
conditions are fully aerobic. Aerobic wort propagation performed at 20±25ëC typically<br />
yields terminal yeast counts of 180±220 10 6 cells/ml. The growth phase requires 24±<br />
48hto complete. Using atwo-tank system, total turn-round times for the brewery phase,<br />
including cleaning <strong>and</strong> filling takes less than seven days. Providing the oxygen supply is<br />
discontinued when growth ceases, the physiology of the yeast remains catabolite<br />
repressed <strong>and</strong> is therefore, similar to that of pitching yeast (Section 12.5.8).<br />
Several propagation systems have been described (Geiger, 1993; Schmidt, 1995;<br />
Br<strong>and</strong>l, 1996; Ashurst, 1990; Boulton <strong>and</strong> Quain, 1999; Westner, 1999). All are provided<br />
withameans ofensuring thathigh ratesofoxygen transfercan beachieved.This requires<br />
two components. Firstly, asystem for delivering sterile oxygen into the growing yeast<br />
culture, usually a sparge ring, c<strong>and</strong>le or similar device <strong>and</strong>, secondly, efficient<br />
mechanical agitation to ensure that oxygen is dispersed throughout the culture. The<br />
essential features are shown in Fig. 13.10. The aspect ratio of the vessel is relatively high<br />
to maximize the path-length for oxygen solution. Vessels are jacketed to facilitate<br />
attemperation by cooling. Either sterile wort is delivered to the wort via an in-line<br />
pasteurizer/heat exchanger, or wort can be sterilized in situ. Sterile oxygen is delivered at<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the base of the vessel via astainless steel c<strong>and</strong>le. High oxygen transfer rates are ensured<br />
by provision of alarge mechanical rouser. If desired, the dissolved oxygen concentration<br />
may be regulated at aset point in afeed-back loop system using output from adissolved<br />
oxygen probe located just under the surface of the liquid. The process generates<br />
considerable foam. Vessels with large freeboards contain this. In addition, foam can be<br />
controlled by application of top pressure <strong>and</strong> reducing agitation <strong>and</strong> gassing rates in<br />
response to atrigger from ahigh-level probe. Propagation in this way is conveniently<br />
performed within two vessels. The first, seed vessel is asimilar to that shown in Fig.<br />
13.10. The operating volume should be around 8hl. The second vessel should be sized to<br />
achieve the desired pitching rate in the first fermentation. For example, apropagation<br />
vessel with an operating volume of 100hl <strong>and</strong> aterminal cell count of 200 10 6 /ml<br />
would generate sufficient yeast to pitch 1300hl of wort at apitching rate of 15 10 6<br />
cells/ml. In contrast to traditional propagation plant large step-up ratios can be used,<br />
typically 1:10 to 1:20.<br />
13.6 Fed-batch cultures<br />
During fermentation or aerobic propagation using wort, yeast metabolism is catabolite<br />
repressed by sugars (Section 12.5.8). This limits the yield of biomass to modest values.<br />
Biomass yields can be dramatically increased by growth under derepressing conditions.<br />
This can be achieved by using an oxidative carbon source such as ethanol or glycerol.<br />
Unfortunately,manyyeaststrainsincludingmostbrewingtypesgrowpoorlyonoxidative<br />
media. Biomass production can be encouraged by fed-batch cultivation, in which the<br />
nutrient medium is supplied to the growing yeast over aperiod of time. The nutrient feed<br />
rate is controlled with respect to growth rate so the sugar concentration remains at alow<br />
derepressing concentration since it is utilized as soon as it comes into contact with the<br />
cells. In <strong>practice</strong>, this can be achieved by inoculating amedium containing all necessary<br />
nutrients but alimited supply of sugar. After inoculation with yeast, growth is allowed to<br />
proceed until the small concentration of sugar is consumed. At this point additional sugar<br />
is added to the culture. Sugar addition is exponential at arate that mirrors the batch<br />
growth curve. Total biomass yields can be increased approximately fivefold using this<br />
procedure, compared to catabolite repressed cultures.<br />
Aerobic fed-batch techniques are used for the production of bakers' yeast. In contrast<br />
to brewing the aim of such processes is to generate biomass. In this respect formation of<br />
ethanol would be wasteful, hence the reliance on this method. Yeast is usually grown on<br />
an undefined medium such as molasses (Barford, 1987). The feedstock can be added in<br />
response to feedback from asensor, which measures the concentration of asugar. In this<br />
case addition rates are regulated to ensure that the sugar concentration remains at zero or<br />
avery low value. Most commercial systems add nutrientat afixed exponential rate based<br />
on empirical observation of the growth of the yeast being cultivated. In order to ensure<br />
that growth remains oxidative it is essential that conditions are aerobic at all times. Since<br />
biomass concentrations reach very high values growth vessels must be capable of very<br />
high rates of oxygen transfer.<br />
Fed-batchpropagationofyeasthasnotbeenappliedtotheproductionofbrewingyeast<br />
although it has been proposed (Masschelein et al., 1994; Naudts et al., 1997).<br />
Theoretically it is an attractive proposition since not only are biomass yields very high<br />
but derepressed yeast contains high concentrations of the essential membrane lipids,<br />
sterols<strong>and</strong>unsaturatedfattyacids(Section12.7).Thesehighlipidlevelsshouldreduceor<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Cooling<br />
section<br />
Wort sterilizer<br />
Sterile inert gas supply<br />
<strong>and</strong> top pressurization<br />
system<br />
Heating<br />
section<br />
Circulation<br />
pump<br />
Cooling<br />
jacket<br />
Temp. <strong>and</strong><br />
DO probes<br />
Wort ring main<br />
Level probes<br />
Variable speed<br />
mechanical<br />
rouser<br />
Working<br />
volume<br />
Internal<br />
baffle<br />
Stainless<br />
steel sinter<br />
Sterile oxygen supply<br />
Pitching<br />
main<br />
Fig. 13.10 Two tank aerobic brewery yeast propagation system. The working capacities of each vessel are 5 brl (8hl) <strong>and</strong> 85 brl (139hl), respectively<br />
(redrawn from Boulton <strong>and</strong> Quain, 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
CIP<br />
Wort in
eliminate the requirement for wort oxygenation. There is areluctance to embrace this<br />
technology. The use of dried yeast in brewing is becoming more common (Fels et al.,<br />
1999). Such yeast is grown using fed-batch cultivation. It seems likely that as approaches<br />
like this see more widespread adoption, so might the underlying growth technique.<br />
13.7 Continuous culture<br />
Batch cultures are closed systems in the sense that exhaustion of a nutrient or<br />
accumulation of agrowth-inhibitory metabolite eventually limits growth. If aconstant<br />
nutrient supply or ameans of removing inhibitory metabolites is provided there is no<br />
reason why growth should not proceed ad infinitum. This is the underlying principle of<br />
open or continuous culture systems. Two approaches are possible (Pirt, 1975). Aplug<br />
flow reactor consists of an extended culture vessel (Fig. 13.11). Acontrolled mixture of<br />
nutrient medium <strong>and</strong> inoculum are introduced into the vessel <strong>and</strong> pumped through it. The<br />
geometryofthevesselisarrangedsuchthatthereisaminimumofmixingofthecontents.<br />
As the medium <strong>and</strong> inoculum pass through the culture vessel the distance travelled is<br />
proportional to the stage in the growth cycle that would have been reached in a<br />
conventional batch cultivation. By manipulation of the flow rate, the dimensions of the<br />
vessel, the composition of the medium, inoculation rate <strong>and</strong> temperature it is possible to<br />
control the composition of the medium issuing from the fermenter. The process can be<br />
made truly continuous by separating the biomass from the process stream exiting the<br />
culture vessel <strong>and</strong> re-circulating a proportion to form the inoculum. Control of the<br />
conditions allow the establishment of asteady state.<br />
Thesecond<strong>and</strong>mostcommonapproachtocontinuousfermentationisthechemostat.In<br />
essence this consists of an attemperated stirred reaction vessel with an inlet medium feed<br />
<strong>and</strong> an outlet for removing product (Fig. 13.12). The pipework is arranged such that the<br />
volume within the reaction vessel remains constant. A variable speed pump controls the<br />
rate of medium inflow <strong>and</strong> product outflow. It is assumed that stirring is perfect such that<br />
incoming medium is instantly <strong>and</strong> homogeneously distributed throughout the growth<br />
vessel. Chemostat cultures are initiated by filling the growth vessel with medium. After<br />
Medium<br />
Stirrer<br />
Reactor – tubular helix<br />
Concentrated biomass return<br />
Continuous<br />
centrifuge<br />
Diverted<br />
product<br />
stream<br />
Product<br />
stream<br />
Fig. 13.11 A plug flow continuous reactor. In the scheme shown, a proportion of the biomass is<br />
recovered by continuous centrifugation <strong>and</strong> fed back into the flow of in-coming medium.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
inoculation,growthisallowedtoproceedforaperiodtogeneratebiomass.Afterthephase<br />
of batch growth the continuous phase is initiated by switching on the medium supply.<br />
The theoretical basis of chemostat operation can be understood starting with a<br />
consideration of the general growth equation, the derivation of which is described in<br />
Section 13.3.<br />
… t†<br />
xt ˆ xoe<br />
Medium<br />
in-flow<br />
Feed pump<br />
Stirred<br />
reactor<br />
Weir outlet<br />
Where x t is the cell population at time t; x o is the cell population at zero time <strong>and</strong> is the<br />
specific growth rate. The latter function is defined as growth rate expressed as a function<br />
of the total biomass concentration. Consequently, it has the units of reciprocal time. The<br />
growth rate is controlled by the same environmental factors that regulate growth in batch<br />
culture such as temperature <strong>and</strong> availability of nutrients. Each organism has a maximum<br />
growth rate ( max) that is determined by the genotype. It is expressed when growth is not<br />
limited by any external influence.<br />
The instantaneous growth rate within a chemostat is described by the equation<br />
13:9<br />
dx<br />
ˆ x 13:10<br />
dt<br />
The rate of loss of cells from the chemostat is given by<br />
Product<br />
out-flow<br />
Fig. 13.12 Simplified representation of a chemostat. The outlet takes the form of a weir. This<br />
ensures that the volume in the reaction vessel remains constant.<br />
dx<br />
ˆ Dx 13:11<br />
dt<br />
D is termed the dilution rate <strong>and</strong> is equal to the flow rate of incoming fresh medium<br />
divided by the volume of the growth vessel. It also has the unit of reciprocal time.<br />
Combining equations 13.10 <strong>and</strong> 13.11 gives an expression showing the net change in the<br />
concentration of biomass at any given time.<br />
dx<br />
ˆ x Dx ˆ x… D† 13:12<br />
dt<br />
From this equation it can be appreciated that for any given dilution rate, the population<br />
density will increase until a constituent of the growth medium becomes limiting <strong>and</strong> its<br />
concentration within the growth vessel tends towards zero. Under these conditions the<br />
biomass concentration <strong>and</strong> composition of the spent medium remain constant <strong>and</strong> are a<br />
function of the dilution rate. This situation is termed the steady state. As the dilution rate<br />
is increased the growth rate ( ) increases to accommodate the increased supply of<br />
nutrients. Eventually, the maximum growth rate ( max) is achieved <strong>and</strong> any further<br />
increase in dilution rate results in a progressive loss of biomass, termed washout.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Both plug flow fermenters <strong>and</strong> especially chemostats have found widespread<br />
application as research tools. Chemostats are very powerful since they provide microbial<br />
biomass with a defined physiological condition. Furthermore, the consequences of<br />
changing physiological state on the biochemistry of the organism by manipulation of the<br />
medium composition, or growth conditions, can be readily assessed. Of greater<br />
significance here, they also form the basis of industrial processes.<br />
Continuous brewing fermentation is an attractive prospect since it provides amethod<br />
of harnessing the power of ahighly active yeast population of defined physiological<br />
condition. Theoretically, it is possible to construct aprocess in which acontinuous infeed<br />
of wort is rapidly transformed into acontinuous outflow of green beer. The yeast is<br />
actively growing at all times, thus the lag phase associated with batch fermentations is<br />
eliminated. The controlled <strong>and</strong> constant conditions within continuous fermenters should<br />
be reflected by beer of consistent composition. Since the process is continuous ipso facto<br />
there is no downtime due to vessel emptying, cleaning <strong>and</strong> re-filling. Similarly, since<br />
yeast is not cropped <strong>and</strong> retained for re-pitching, yeast h<strong>and</strong>ling is much simplified.<br />
In <strong>practice</strong>, there are also many disadvantages to continuous fermentation. The flow<br />
ratescanbevariedwithinonlycomparativelynarrowlimits<strong>and</strong>onlyasinglebeerquality<br />
can be produced at any given time. Consequently, continuous systems are inflexible.<br />
There is aneed for aconstant supply of wort. Most breweries have plant suitable for<br />
discontinuous wort production (Chapter 6) <strong>and</strong> therefore a storage facility must be<br />
installed. The risks of microbial spoilage with bulk wort storage are considerable. The<br />
risks of contamination extend to the continuous fermenter. The consequences of<br />
contamination in the reactor are grave since start-up times <strong>and</strong> establishment of steadystate<br />
conditions are long. Some brewing yeast strains are genetically unstable (Section<br />
11.8.2). Selection of mutant strains in continuous fermenters can have disastrous<br />
consequences.Forexample,someproductionscalecontinuousprocessesrelyontheyeast<br />
being flocculent for retention within the vessel. Selection of non-flocculent variants<br />
results in washout.<br />
The nature of brewery primary fermentation precludes the use of achemostat. The<br />
assimilation of sugars by yeast growing on wort is an ordered process. The presence of<br />
glucose inhibits the assimilation of maltose (Section 12.4.1). Furthermore, the<br />
production of abalanced spectrum of flavour compounds is dependent on controlled<br />
yeast growth. In achemostat culture, yeast growth extent may be significantly different<br />
from that seen in batch fermentation. Similarly, the constant addition of fresh wort can<br />
result in abnormal sugar utilization. For these reasons, continuous primary fermentation<br />
is best performed using aplug flow type reactor. Unfortunately, it is impossible entirely<br />
to eliminate back mixing. To avoid these problems <strong>and</strong> by way of compromise, multistage<br />
continuous systems may be used. In this case the physical separation of the process<br />
liquid into discrete reaction vessels allows the essential characteristics of abatch culture<br />
to be incorporated into asemi-continuous process (Stratton et al., 1994; Williams <strong>and</strong><br />
Ramsden, 1963). Alternatively, the fermenter takes the form of avertical column (tower)<br />
in which wort is introduced at the base (Seddon, 1975). Use of aflocculent strain not<br />
only retains yeast within the vessel but also provides a self-generating gradient of<br />
biomass through which the wort passes. The pinnacle of interest in continuous<br />
fermentation was reached during the 1960s. During this period many systems were<br />
developed, some of which were used at production scale. Unfortunately, their use was<br />
bedevilled by failures. With afew notable exceptions the use of continuous fermentation<br />
systems was discontinued. Since that time there has been a resurgence of interest,<br />
especially using immobilized yeast reactors, as described in Section 13.8. Afull review<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
of the historical aspects <strong>and</strong> current status of continuous fermentation can be found in<br />
Boulton <strong>and</strong> Quain (2001).<br />
13.8 Immobilized yeast reactors<br />
Immobilized yeast reactors are refinements of continuous fermentation systems. In the<br />
latter, some of the yeast exits from the vessel with the spent medium. This necessitates<br />
the use of additional plant, usually a continuous centrifuge, for separation of yeast <strong>and</strong><br />
product. In continuous immobilized yeast reactors the yeast is retained within the vessel.<br />
Fresh medium enters the reactor <strong>and</strong> passes through the yeast biomass where it is<br />
transformed into product. The latter exits from the reactor essentially free from yeast.<br />
Immobilized systems have several advantages compared to conventional continuous<br />
reactors. Process times are rapid <strong>and</strong> process efficiencies are high because of the<br />
combination of elevated biomass concentration <strong>and</strong> high volume throughput. Thus,<br />
because yeast loss is restricted, it is possible to use flow rates that would cause washout in<br />
a conventional chemostat. In an immobilized yeast reactor the relation between dilution<br />
rate <strong>and</strong> biomass concentration does not hold since cells are retained. Hence, it is possible<br />
to have high dilution rates <strong>and</strong> high biomass concentrations. High biomass concentration<br />
restricts growth, which also engenders high process efficiencies. High productivity allows<br />
the use of comparatively small immobilized yeast reactors. Clarification of the product<br />
stream is simplified because of the retention of biomass. In essence, the yeast in an<br />
immobilized system functions as a biocatalyst <strong>and</strong> no actual growth is necessary. The<br />
process has been defined as cells physically confined or localized within a specific region<br />
of space with retention of their catalytic activity, if possible or even necessary their<br />
viability, which can be used repeatedly <strong>and</strong> continuously (McMurrough, 1995).<br />
Several systems for immobilization are used (McMurrough, 1995). Retention of<br />
flocculent yeast within a tower fermenter using upward flow is the simplest method. It is<br />
of limited use since it can be used only with very flocculent strains <strong>and</strong> high flow rates<br />
eventually result in washout of biomass. There are four methods for true immobilization<br />
of yeast. These are retention by a semi-permeable membrane, attachment to a surface,<br />
entrapment within a porous polymer <strong>and</strong> colonization of a porous material. Membrane<br />
reactors suffer the major disadvantage that the rate of exchange of solutes is slow. In<br />
consequence, they have not found application in production scale reactors.<br />
Entrapment of yeast within the matrix of a porous polymer is possibly the most widely<br />
used method of immobilization. Polymers that have been used include polyacrylamide,<br />
calcium alginate, -carrageenan, agarose, pectin, chitin <strong>and</strong> gelatin (Godia et al., 1987).<br />
Polymeric supports are conveniently formed into beads, typically around 0.3 mm diameter.<br />
Beads of this dimension offer the best performance with regard to diffusion of nutrients <strong>and</strong><br />
metabolic by products from the surrounding medium to the immobilized cells. The<br />
mechanical strength of the bead is a function of the degree of cross-linking of the polymer.<br />
The latter also controls diffusivity of nutrients <strong>and</strong> metabolites from the medium to the<br />
yeast cells. The latter parameter is vital to the efficiency of these bioreactors hence it is<br />
usual to sacrifice some mechanical strength in the interests of productivity. Since the beads<br />
are also compressible, entrapped yeast bioreactors are usually operated with the process<br />
flow directed in a vertically upward direction. Beads with very low mechanical strength can<br />
be disrupted by the evolution of gas bubbles <strong>and</strong> retention of yeast is comparatively poor.<br />
Yeast cells can be attached to inert surfaces such as wood chips, ceramics, glass,<br />
cellulose, stainless steel <strong>and</strong> various resins (Godia et al., 1987; Ryder et al., 1995). The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mechanism of attachment is via acombination of electrostatic <strong>and</strong> hydrophobic binding<br />
(Mozes et al., 1987). For brewing applications the use of DEAE cellulose has found<br />
favour. This material is described as granulated derivatized cellulose (GDC) <strong>and</strong> is sold<br />
under the trade name Spezyme Õ (Cultor Finl<strong>and</strong>). The beads have adiameter of 0.4<br />
0.8mm <strong>and</strong> are capable of bearing ayeast loading of 500 10 6 cells per gwet weight of<br />
carrier. Following prolonged use the beads can be regenerated by treatment with NaOH<br />
(2% w/v) at 80ëC (Pajunen, 1995).<br />
Colonization of porous materials is acombination of entrapment <strong>and</strong> surface binding.<br />
Several types of support have been used, the most promising of which are glass beads<br />
(SIRAN Õ ,Schott Engineering Company, Mainz, Germany) <strong>and</strong> ceramic rods containing<br />
amatrix of silicon carbide. The former consists of beads with adiameter of 1 3mm<br />
prepared from amixture of glass powder <strong>and</strong> salt. After the beads are manufactured the<br />
salt is dissolved to leave pores of 60 300 mdiameter. The proportion of each bead<br />
which is pore accounts for approximately 60% of the total. This gives biomass loadings<br />
of approximately 15 10 6 cells per gbead (Breitenbucher <strong>and</strong> Mistler, 1995). Ceramic<br />
cylinder supports have been described by Krikilion et al., (1995). Each cylinder element<br />
contains channels through which the process liquid flows. The channels pass through a<br />
poroussiliconcarbidematrix.Thesurfaceporesofthematrixareapproximately8±30 m<br />
in diameter <strong>and</strong> internal pores have adiameter of 100±150 m. The small outer pores<br />
allow entry of yeast cells, which then grow <strong>and</strong> colonize the comparatively larger lumen.<br />
Both glass<strong>and</strong> ceramicsupports are incompressible <strong>and</strong> so there islittle restriction on the<br />
flow rates that can be used. They can be used in reactors with upward or downward flow.<br />
Similarly, the robustness of the materials protects then from damage due to CO2<br />
evolution. On the other h<strong>and</strong>, because cells penetrate into the supports, they are more<br />
difficult to clean for regeneration purposes.<br />
All immobilized reactor systems have afinite life, which is limited by the gradual<br />
build-up of debris <strong>and</strong> deterioration of the support <strong>and</strong> the yeast population. The duration<br />
of the lifespan <strong>and</strong> the ease <strong>and</strong> cost of regeneration are important considerations when<br />
making achoice of support <strong>and</strong> reactor. Comparatively expensive support materials such<br />
as glass beads <strong>and</strong> ceramics require treatment with hot NaOH or an oxidizing agent such<br />
as hydrogen peroxide, followed by steam sterilization. Relatively inexpensive supports<br />
such as alginate beads or wood chips can be discarded after asingle use.<br />
Immobilization influences yeast physiology. From the st<strong>and</strong>point of the brewing<br />
process some of the effects are beneficial others are not. Several reports describe<br />
enhanced rates of glycolysis <strong>and</strong> ethanol formation in immobilized cells compared to<br />
freely suspended cells (Galazzo <strong>and</strong> Bailey, 1989, 1990; Aires Barros et al., 1987). In<br />
contrast, rates of growth, as measured by biomass increase are reduced in the case of<br />
immobilized cells. In one report (Aires Barros et al., 1987) the rate of ethanol formation<br />
in an immobilized system was 50% greater than that of the same concentration of freely<br />
suspended cells, but the biomass yield was reduced by 30% <strong>and</strong> there were increases in<br />
trehalose <strong>and</strong>glycogen reservesin immobilized cells. These observations suggest thatthe<br />
immobilized yeast cells enjoy aprotected environment in agreement with the finding that<br />
they exhibit an increased tolerance to ethanol (Holcberg <strong>and</strong> Margalith, 1981). This is in<br />
accord with the assertion that micro-organisms in nature usually grow in association on<br />
surfaces, as in biofilms (Section 17.7.1).<br />
Adverse effects on yeast due to immobilization are probably related to the effects of<br />
restricted transfer of solutes. Immobilized systems involving colonization of surfaces or<br />
in polymeric beads provide a continuum of environments from surface to core. Cells that<br />
are buried deep in the matrix of the bead commonly exhibit morphological abnormalities<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
such as pleomorphism, increased ploidy <strong>and</strong> failure of daughter cells to separate (Aires<br />
Barros et al., 1987; Koshcheyenko et al., 1983). These cells are likely to be subjected to<br />
very hostile conditions, such as starvation, anaerobiosis, reduced water activity, high<br />
ethanol concentration <strong>and</strong>high dissolvedCO 2.Attachmenttosurfacesmay, insomeway,<br />
disrupts the cell cycle. With carriers that rely on binding of cells to surfaces,<br />
morphological abnormalities are less common. This suggests that mass transfer has the<br />
greatest influence. Whatever the precise reason it explains the low growth rate that is<br />
characteristic of immobilized systems.<br />
The fact that immobilized systems function as biocatalysts where there is very limited<br />
growth is both astrength <strong>and</strong> aweakness. They are most suitable for use in processes<br />
where limited metabolism is required. Thus, they have been widely used for the<br />
production of low-alcohol beers by limited fermentation <strong>and</strong> for continuous diacetyl<br />
removal. In the first instance, ethanol formation can be discouraged by the use of a<br />
combination of low temperature <strong>and</strong> anaerobiosis (Breitenbucher <strong>and</strong> Mistler, 1995; van<br />
de Winkel et al., 1991). Debourg et al., (1994) demonstrated that immobilized yeast was<br />
as effective at removing wort carbonyls as freely suspended cells. Diacetyl removal is<br />
accomplished by passing green beer, ex-primary fermenter through areactor containing<br />
immobilized yeast (Pajunen <strong>and</strong> Gronquist, 1994). The high concentration of yeast<br />
removes diacetyl very rapidly. It is necessary to heat green beer (c. 10min. at 90ëC;<br />
194ëF) prior to passage through the bioreactor to ensure that all precursor -acetolactate<br />
is converted to diacetyl.<br />
The strictures that apply to the application of continuous processes to primary<br />
fermentation also hold for immobilized systems. Limited yeast growth <strong>and</strong> the need for<br />
ordered assimilation of metabolites to produce balanced quantities of flavour metabolites<br />
are problematic. Inmostcases the optionof using series ofmultipletanksofimmobilized<br />
yeast has been adopted in order to physically separate the process stages occurring in<br />
conventional batch fermentation. A fuller description of prototype <strong>and</strong> commercial<br />
immobilized yeast reactors used in brewing may be found in Boulton <strong>and</strong> Quain (2001).<br />
13.9 Growth on solid media<br />
In the majority of immobilized yeast systems the cells grow attached to inert surfaces. In<br />
this situation the cells obtain nutrients from the surrounding liquid medium <strong>and</strong> the<br />
surface merely forms asubstrate for attachment. An alternative growth habit is that of<br />
yeastgrowingonthesurfaceofnutrientmediasolidifiedwitheitheragarorgelatin.Yeast<br />
cells multiply <strong>and</strong> form masses, termed colonies, on the surface of the medium. Colonies<br />
are roughly circular viewed from above <strong>and</strong> dome shaped in section. Colony sizes range<br />
from 1 10 7 cells to more than 10 9 cells depending on the number per plate. The precise<br />
shapes of the colonies are often characteristic of individual strains <strong>and</strong> are used as an aid<br />
to identification. In brewing, the shape <strong>and</strong> colours of colonies that form on generalpurpose<br />
media such as Wallerstein Laboratory nutrient agar, can be used to check strain<br />
purity (Section 17.3.6). As discussed later, the giant colony technique is atraditional<br />
method for brewing yeast strain differentiation.<br />
The pattern of growth on solid media by yeast is a function of the genotype of the<br />
particular strain <strong>and</strong> how the cells respond to the available nutrients. The kinetics of<br />
colonial growth has been studied by Kamath <strong>and</strong> Bungay (1988). The colony increases in<br />
size at the periphery <strong>and</strong> in height. The rate of peripheral growth is linear <strong>and</strong> not<br />
exponential. It can be expressed by the following equation:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
t ˆKrt ‡r0<br />
13:13<br />
Where, rt is the linear radial growth rate, Krt is the radial growth rate constant, r0 is the<br />
radius at zero time, ris the colony radius at time t.<br />
In the giant colony technique, yeast is grown on wort medium solidified with gelatin.<br />
Plates are incubated for 3±6 weeks at temperatures of 15±18ëC; 59±69.4ëF (Hall, 1954).<br />
Several characteristics of colonies may be of diagnostic significance. Colonies are<br />
described asbeingmattorshiny.Theprofiles ofcoloniesviewedinsection<strong>and</strong>fromabove<br />
may be simple or very complex. Variations in the peripheral margins of colonies include<br />
smooth circular, fringed, irregular <strong>and</strong> lobate. The surfaces of colonies range from smooth<br />
to concentric or radial striated <strong>and</strong> deep radial valleyed types. Variations in colonial<br />
profiles include convex, flat, with acentral dome, wrinkled <strong>and</strong> crateriform. The variety of<br />
colonial morphologiesimpliesthatthemassofcellsisheterogeneous. Thisispredictable in<br />
that cells in the growing colony do not all have equal access to nutrients in the medium <strong>and</strong><br />
to oxygen. The variety of responses to these inequalities in nutrient supply explains the<br />
characteristics of colonial morphology. For example, the formation of afringe around the<br />
margin of a colony is a dimorphic response in which cells at the margins adopt a<br />
pseudomycelial form. This may represent amechanism by which peripheral cells extend<br />
the area over which they are able to assimilate nutrients. Other shapes represent the<br />
response ofindividualcellswithinthedevelopingcolony togreaterorlesser concentrations<br />
of nutrients <strong>and</strong> metabolic by-products. Colonies are highly organized structures. Agene,<br />
given the name IRRI, has been isolated that has no known role in cells growing suspended<br />
in liquid media but is required for colony formation on solid substrates (Kurlanzka et al.,<br />
1999). As with the case of biofilms it is highly likely that colonies represent astrategy in<br />
which the population ensures survival by co-operative behaviour.<br />
13.10 Yeast identification<br />
The taxonomy of yeast, including brewing strains is discussed in Chapter 11 (Section<br />
11.2). The methods used for the cultivation of yeast for the purposes of identifying<br />
contaminants are described in Chapter 17 (Sections 17.3.5; 17.3.6). Here the methods<br />
used for the differentiation of brewing yeasts are described. Methods are roughly<br />
divisible into two groups, the traditional <strong>and</strong> the modern. Traditional methods are often<br />
based on conventional microbiological techniques. They rarely have taxonomic<br />
significance but have evolved within individual breweries to meet the need for<br />
identification <strong>and</strong> differentiation of proprietary brewing yeast strains. Modern techniques<br />
frequently have taxonomic significance. Usually they have been developed within<br />
academic research organizations <strong>and</strong> are not specifically targeted at brewing. Commonly<br />
they use sophisticated <strong>and</strong> costly apparatus. Traditional approaches are still used to avoid<br />
the cost <strong>and</strong> requirement for skilled operatives needed for many modern techniques.<br />
Modern methods may be used in confirmatory tests by third parties. However, the use of<br />
molecular genetic analyses is becoming more commonplace <strong>and</strong> inexpensive. These may<br />
soon be seen as the methods of choice for routine brewery QA testing.<br />
13.10.1 Microbiological tests<br />
Plating microbiological samples on solid nutrient media is routinely used for enumerating<br />
microbial populations. By using selective or chromogenic media the method can have<br />
diagnostic significance. The method most widely used for differentiating brewing from<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
wild strains relies on the ability of the latter to grow in the presence of inhibitors such as<br />
Cu 2+ orthe antibiotic cycloheximide (Section 17.3.6). General-purpose media such as<br />
WLN help in the differentiation of brewing strains. WLN contains the indicator dye<br />
bromocresol green <strong>and</strong> this causes yeast colonies to take on avariety of shades of green.<br />
The particular shade of colouring is characteristic of some but not all strains. The plates<br />
mustnotoverloaded.Themethodneedsaskilledpractitionerwithexperienceoftheyeast<br />
strains being examined. The use ofgiant colony morphology isdescribed in Section 13.8.<br />
The requirement for several weeks' incubation reduces its usefulness.<br />
Differentiation of lager <strong>and</strong> ale colonies can be accomplished by plating out onto a<br />
medium containing 5-bromo-4-chloro-3-indoyl- -D-galactoside, known as X- -Gal<br />
(Tubb <strong>and</strong> LiljestroÈm, 1986). Lager, but not ale, strains contain -galactosidase which<br />
cleaves the chromogenic substrate to give a product which forms a blue precipitate.<br />
Colonies of ale yeasts remain colourless whereas lager types assume a blue colour.<br />
13.10.2 Biochemical tests<br />
Several tests have been developed, which probe the biochemical properties of individual<br />
yeast strains. Many of these tests are part of the st<strong>and</strong>ard armoury of the microbiologist.<br />
There is some overlap with the methods discussed in the previous section. Other<br />
techniques probe aspects of biochemistry that are more directly relevant to brewing<br />
performance.<br />
The ability of an organism to utilize selected carbon sources under aerobic conditions<br />
(assimilation) or anaerobic conditions (fermentation) has been widely used as a method<br />
of differentiation <strong>and</strong> identification. Several techniques based on this approach are used.<br />
They share in common the assessment of growth on a basal nutrient media supplemented<br />
with a variety of carbon sources. Basal media include peptone water <strong>and</strong> yeast nitrogen<br />
base (Section 17.3.6). Liquid media require the use of pure cultures, frequently indicator<br />
dyes are included <strong>and</strong> a trap for detecting any CO2 formed. Solid media are also usually<br />
inoculated with pure cultures. However, with replica plating (Section 17.3.6) it is possible<br />
to assess mixed populations. Commercial kits are available for identifying particular<br />
groups of micro-organisms, both bacteria <strong>and</strong> yeast (for example, API test kits,<br />
bioMeÂrieux, Marcy-l'Etoile, France). These take the form of strips containing wells each<br />
of which is filled with a basal medium <strong>and</strong> a supplement of a different carbon source.<br />
After inoculation <strong>and</strong> incubation, growth is indicated by the change in colour of an<br />
indicator dye. As with many of the general microbiological approaches, these methods<br />
are best suited to identifying wild yeast strains. One assimilation test does have value.<br />
Since lager strains possess -galactosidase <strong>and</strong> ale strains do not, the former are able to<br />
hydrolyse melibiose <strong>and</strong> grow on the released sucrose <strong>and</strong> galactose.<br />
Strain identification can be based on assessment of performance in laboratory wort<br />
fermentations. This approach is of obvious value since it assesses the properties that will<br />
be exhibited during use in the brewery. In this respect, many of the tests are of the<br />
trueness-to-type variety. Whilst these may not have taxonomic significance they are a<br />
useful means of checking for strain drift. Several pieces of equipment have been designed<br />
for carrying out laboratory fermentations. These range from small (100 ml) stirred<br />
hypovials (Quain et al., 1985) through to larger stirred fermenters with facilities for<br />
controlling headspace gas. The most commonly used piece of apparatus is the EBC tall<br />
tube. These are attemperated glass or stainless steel tubes with a capacity of two litres <strong>and</strong><br />
a high aspect ratio (150 cm 5 cm diameter) supposedly reflective of a fermenter. They<br />
are usually used in banks of several tall tubes to permit simultaneous fermentations.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Automated versions are available, some of which are fitted with devices for automated<br />
sampling <strong>and</strong> data acquisition (Sigsgaard <strong>and</strong> Rasmussen, 1985; Sk<strong>and</strong>s, 1997).<br />
Fermentation performance in tall tubes is assessed by five criteria; formation of ayeast<br />
head, formation of a yeast sediment, attenuation rate, degree of attenuation <strong>and</strong><br />
clarification after fining.<br />
Laboratory fermentations can be used to assess other growth-related properties of<br />
yeast. For example, the effect of varying temperature. Thus, ale strains on average have<br />
higher maximum growth temperatures (37 40ëC; 98.6 104ëF) compared with lager<br />
strains (31.5 34ëC; 88.7 93.2ëF) (Walsh <strong>and</strong> Martin, 1977). The oxygen requirement<br />
needed for satisfactory fermentation performance varies between individual yeast strains<br />
(Section 12.6). Pitching aliquots of wort containing varying dissolved oxygen<br />
concentration then observing subsequent fermentation performance can assess this<br />
property.<br />
13.10.3 Tests based on cell surface properties<br />
The surface properties of yeast cells underpin two types of differentiating tests. These are<br />
firstly, those that assess the immunological properties of cells <strong>and</strong>, secondly, those that<br />
measure flocculation.<br />
Immunological analyses rely on interactions between antigens (the yeast cell or<br />
fraction thereof) <strong>and</strong> antibodies. The latter are obtained by prior inoculation of the<br />
antigen fraction into asuitable mammalian host, usually arabbit. Antibodies are raised in<br />
response to the antigen <strong>and</strong> these can be isolated <strong>and</strong> purified. Components of the yeast<br />
cell wall, principally mannan side chains, have antigenic activity (Section 11.6.1).<br />
Reactions between antigens <strong>and</strong> antibodies are specific. The complexes formed can be<br />
visualized by coupling a fluorescent molecule to the antibody. A much-used method is<br />
the ELISA technique (enzyme-linked immuno-absorbent assay), in which a membrane is<br />
coated with a layer of antibody, which is reactive with the antigen of interest. The antigen<br />
mixture is added to the membrane. The membrane is then washed, which removes all but<br />
the bound antigen. The complex is then treated with a second batch of antibody, the same<br />
as that attached to the membrane, although in this case it is conjugated with an enzyme.<br />
The second antibody fraction binds to the antigen-antibody complex. After a second wash<br />
to remove unbound antibody, a substrate is added, which is acted upon by the enzyme <strong>and</strong><br />
in doing brings about a colour change. This colour change allows visualization <strong>and</strong><br />
quantification of bound antigen.<br />
Methods of assessing yeast flocculence have a long history. They cannot be used to<br />
provide a positive identification but are useful nonetheless for assurance that yeast strains<br />
are behaving normally. Four methods have gained acceptance within the brewing industry.<br />
These are the procedures of Burns (Burns, 1941), Gillil<strong>and</strong> (1951), Hough (1957) <strong>and</strong><br />
Helms (Helms et al., 1953). Of these only the Gillil<strong>and</strong> <strong>and</strong> Helms methods remain in<br />
common use. The Hough method tests for flocculation of aliquots of washed cells<br />
suspended in a solution of calcium chloride adjusted to pH 3.5 or pH 5.0. Yeast strains are<br />
classified based on whether flocculation occurs under these conditions. If yes, does<br />
addition of maltose disperse flocs? If no, does addition of ethanol result in flocculation? If<br />
no, does the addition of addition of a second yeast strain (NCYC 1108) result in coflocculation.<br />
In the Helms method a sample of pitching yeast is washed by suspension <strong>and</strong><br />
centrifugation in a solution of calcium sulphate. The washed yeast is re-suspended in<br />
calcium sulphate adjusted to pH 4.5 in a graduated centrifuge tube. After incubation at<br />
20 ëC for 20 minutes flocculation is assessed based on the volume of sediment formed.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Both the Burns <strong>and</strong> Gillil<strong>and</strong> procedures assess the occurrence of flocculation after<br />
yeast is recovered from laboratory wort fermentations. In the Burns procedure, the extent<br />
of floc formation is recorded when the yeast is suspended in beer, distilled water <strong>and</strong><br />
acetate buffer, pH 4.6. In the Gillil<strong>and</strong> method yeast is grown on agar medium to obtain<br />
separate colonies. Fifty colonies are picked off the plate <strong>and</strong> each is inoculated into 5 ml<br />
trub-free hopped wort. After three days static incubation, at 27 ëC, each culture is<br />
examined for the formation of sediment. After pouring off all but 0.5 ml of the liquid the<br />
yeast is gently agitated <strong>and</strong> examined. Gillil<strong>and</strong> classified yeast into four groups on the<br />
basis of their flocculence characteristics. Class I types were completely dispersed<br />
throughout. Class II yeast sediment towards the end of the incubation. The sediment is<br />
granular in appearance when re-suspended. Class III types behave similarly to the Class<br />
II but the sediment is more difficult to re-suspend <strong>and</strong> forms large flakes. Class IV yeast<br />
sediments very early in the growth phase. The sediment re-suspends to form loose<br />
flakes.<br />
13.10.4 Non-traditional methods<br />
Several methods for differentiating yeast strains are relatively new at the time of writing.<br />
They fall into two broad groups. Firstly, those which are based on an analysis of cell<br />
composition <strong>and</strong>, secondly, those which probe the genome of the cell. Methods based on<br />
cellular composition rely on whole cell analyses, for example, pyrolysis mass<br />
spectrometry <strong>and</strong> Fourier transform infra-red spectroscopy. Alternatively, specific subcellular<br />
fractions can be extracted <strong>and</strong> analysed, for example proteins <strong>and</strong> lipids. Methods<br />
for strain differentiation based on cell composition require that the yeast has been<br />
cultivated under defined conditions in order to eliminate differences due to variations in<br />
physiological state. Genetic analyses have the advantage that the composition of the<br />
genome is relatively constant <strong>and</strong> independent of physiological state.<br />
Pyrolysis gas chromatography <strong>and</strong> pyrolysis mass spectrometry rely on heating a<br />
sample of biomass in an inert atmosphere to around 550 ëC (1022 ëF). This causes the<br />
cells to decompose into a mixture of low molecular weight volatile fragments (Goodacre,<br />
1994). The fragments are separated using gas chromatography or the more powerful<br />
technique of gas chromatography mass spectrometry (GC-MS). The pattern of fragments<br />
that are generated are characteristic for individual or closely related groups of strains<br />
(Timmins et al., 1998). Fourier transform infra-red spectroscopy provides a fingerprint of<br />
whole cells by measuring the interactions between infra-red radiation <strong>and</strong> intracellular<br />
components such as nucleic acids, proteins, membranes <strong>and</strong> cell wall polysaccharides.<br />
For microbial cells, the mid-IR range (4000 400 cm 1 ) provides the best resolving<br />
power. The method is capable of distinguishing bacteria at the strain level (Helm et al.,<br />
1991). The method has been successfully used for differentiating brewing yeast strains<br />
(Timmins et al., 1998). The proteome of cells can be extracted <strong>and</strong> separated by<br />
electrophoresis <strong>and</strong> visualized using techniques such as Western blotting. Based on<br />
differences in genotype it would be predicted that proteome analyses would be different<br />
for individual strains <strong>and</strong> therefore of taxonomic significance. This has been verified in a<br />
study of 29 enological strains of S. cerevisiae (van Vuuren <strong>and</strong> van der Meer, 1987).<br />
Total fatty acids can be extracted from yeast using a solvent such as a mixture of<br />
chloroform <strong>and</strong> methanol. In the form of methyl esters, the mixture of fatty acids can be<br />
separated using capillary gas liquid chromatography. The spectrum of fatty acid methyl<br />
esters <strong>and</strong> their relative abundance has been used to differentiate 13 strains of S.<br />
cerevisiae (Augustyn <strong>and</strong> Kock, 1989).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The most precise methods for strain differentiation are those based on analysis of the<br />
nucleic acids of the genome. These methods generate so-called genetic fingerprints<br />
(Chapter 11; Section 11.8.1). Yeast strains can be positively identified <strong>and</strong> differentiated<br />
using restriction fragment polymorphism (Schofield et al., 1995), polymerase chain<br />
reaction (de Barros et al., 1998) <strong>and</strong> karyotyping (Casey, 1996).<br />
13.11 Measurement of viability<br />
Viability of micro-organisms is usually defined as the ability to reproduce. This is a<br />
useful parameter to determine since it provides apredictive measure of the ability of<br />
yeast to reproduce when pitched into wort. In traditional microbiology, viability is<br />
assessed by plating out serial dilutions of suspension of cells onto asuitable medium.<br />
Each colony that develops after incubation is assumed to have derived from asingle cell,<br />
therefore the colony count is directly related to the number of viable cells in the original<br />
suspension. Using this approach the viable cell concentration is usually expressed as the<br />
number of colony forming units (cfu). The colony counting procedure takes several days<br />
to obtain aresult. There is arapid version of the test in which micro-colonies are allowed<br />
to grow on nutrient agar poured into awell in aspecially designed microscope slide. The<br />
micro-colonies can be counted using amicroscope. This reduces the time taken to obtain<br />
a result from days to several hours. Even using the rapid method, colony-counting<br />
techniques aretooslowtobeusedtosupportdecisionsregardingthefitnessofyeasttobe<br />
pitched. Rapid methods for the determination rely on the use of vital stains. These are<br />
dyes which are excluded from viable cells but not dead ones. Alternatively, they may be<br />
taken up by all cells <strong>and</strong> subject to metabolic modification by living cells. The latter is<br />
accompanied by acolour change. Several dyes have been used to assess viability. Some<br />
of these, together with their mode of action are shown in Table 13.2.<br />
The most used method for the assessment of yeast viability in breweries employs<br />
methylene blue <strong>and</strong> ahaemocytometer counting chamber (EBC Analytica Microbiologica,<br />
1992). The haemocytometer consists of aglass microscope slide, which contains<br />
two chambers of known volume. The bases of the chambers are divided into anumber of<br />
small squares to form agrid which facilitates counting. To determine viability asuitable<br />
dilution of yeast slurry is mixed with an equal volume of a solution of methylene blue <strong>and</strong><br />
placed into the haemocytometer. Viable cells take up methylene blue <strong>and</strong> reduce the dye<br />
to the colourless leuco form. Dead cells cannot reduce the dye <strong>and</strong> stain blue. The relative<br />
proportions of colourless <strong>and</strong> blue cells are counted <strong>and</strong> by calculation the total <strong>and</strong><br />
viable yeast count in the slurry is determined.<br />
Compared to plate counts, the methylene blue method overestimates viability. While<br />
at values greater than 90% agreement is reasonable, the disparity becomes increasingly<br />
marked the lower the true viability (Parkinnen et al, 1976). It has been suggested that<br />
other dyes are more reliable than methylene blue. In one study, methylene blue (with<br />
Safranin O counterstain), citrate methylene blue, alkaline methylene blue, citrate<br />
methylene violet <strong>and</strong> alkaline methylene violet were compared with plate count<br />
determinations of viability (Smart et al., 1999). A variety of strains were tested in various<br />
physiological conditions ranging from exponential phase, through stationary phase,<br />
starved <strong>and</strong> heat-killed. It was concluded that citrate methylene violet produced the most<br />
reliable results. Nevertheless, methylene blue continues to be used widely. Providing the<br />
analysis is performed by a skilled operator <strong>and</strong> the viability of the yeast is greater than c.<br />
80% it provides a reliable result. Counting cells stained with dyes can be automated to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 13.2 Methods for measuring yeast viability<br />
Method Mechanism Ref.<br />
Methylene blue staining<br />
Viable cells reduce dye to colourless form, dead cells stain blue<br />
[1]<br />
Methylene blue with Safranin O counterstain<br />
Fluorescein diacetate<br />
Bis(1,3-dibutylbarbituric acid trimethine<br />
oxonol) [DiBAC4]<br />
Propidium iodide (+ fluorescein diacetate)<br />
ChemChrome Y<br />
Mg 1-aniline-8-naphthalene sulphonic acid<br />
(Mg-ANS)<br />
2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)<br />
amino]-2-deoxyglucose (2NBDG)<br />
Rhodamine 123<br />
Counterstain reportedly improves contrast between live <strong>and</strong> dead cells <strong>and</strong> possibly differentially stains<br />
stressed cells<br />
Esterases in viable cells cleave molecule to release fluorophor, fluorescein<br />
Anionic dye excluded by viable cells, non-viable cells are fluorescent<br />
Excluded by live cells, stains dead cells red by binding to DNA. With counterstain, viable cells are<br />
fluorescent green<br />
Taken up by viable cells <strong>and</strong> cleaved enzymically to release fluorophore<br />
Fluorophore only taken up by viable cells where it binds to proteins<br />
Fluorescent derivative of glucose only taken up by viable cells<br />
Cationic fluorophore taken up by viable cells with functional oxidative mitochondria. Not useful for<br />
viability measurement of anaerobic repressed yeast<br />
[1] IOB Methods of Analysis, 1997; [2] Jones, 1987; [3] Chilver et al., 1978; [4] Dinsdale et al. 1999; [5] Lloyd <strong>and</strong> Hayes, 1995; [6] Deere et al., 1998; [7] McCaig, 1990; [8] Oh <strong>and</strong><br />
Matsuoka, 2002; [9] Dinsdale <strong>and</strong> Lloyd, 1995.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
[2]<br />
[3]<br />
[4]<br />
[5]<br />
[6]<br />
[7]<br />
[8]<br />
[9]
emove human error, for example, with electronic image analysis (Raynal et al., 1994).<br />
The radiofrequency permittivity biomass meter described in Section 13.2 is responsive<br />
only to the viable fraction of yeast slurries. Since it does not respond to the non-viable<br />
fraction it does not provide a measure of viability. A prototype modified laboratory<br />
version is described in (Boulton et al., 2001) which uses a combination of radiofrequency<br />
permittivity <strong>and</strong> a fluorescent vitality stain to detect both viable <strong>and</strong> non-viable yeast<br />
cells. Should this become available commercially it would allow automatic measurement<br />
of viability.<br />
Flow cytometry has been used for viability determinations. This apparatus forces a<br />
suspension of yeast through a small nozzle such that the cells form a single-file stream.<br />
The cells are counted by passage through a laser beam. In addition, devices of various<br />
types are provided for the detection of cells stained with specific dyes, most usually<br />
fluorescent types such as fluorescein diacetate (Lloyd, 1993; Bouix <strong>and</strong> Leveau, 2001).<br />
The procedure gave a good correlation between fluorescent staining <strong>and</strong> a plate counting<br />
technique. Unfortunately, flow cytometers are costly <strong>and</strong> not likely to find routine use in<br />
any but the most lavishly appointed routine quality assurance laboratories. However, the<br />
instruments are capable of much more than simple measurement of viability as discussed<br />
in the following section.<br />
13.12 Assessment of yeast physiological state<br />
The results of viability tests are used for the calculation of the quantity of yeast to be<br />
added to wort to achieve a desired pitching rate. They are also used to assess the quality<br />
of yeast. Typically, an arbitrary value is chosen, usually around 90%, below which the<br />
yeast is considered unfit for use. The assumption is made that if yeast viability is low then<br />
the viable fraction is probably stressed. In recent years, a number of additional tests have<br />
been proposed that seek to probe the physiological state of the viable fraction of yeast<br />
populations. These are `vitality tests' (Lentini, 1993). The rationale behind the need for<br />
these tests is that current fermentation <strong>practice</strong> exposes yeast to a plethora of influences,<br />
which together have the potential to influence yeast physiology in perhaps unexpected<br />
ways. These influences have been comprehensively reviewed by Heggart et al. (1999).<br />
They include environmental effects such as osmotic stress, barometric stress, oxidative<br />
stress, mechanical stress, pH effects <strong>and</strong> temperature. These are coupled with genetic<br />
effects such as mutation <strong>and</strong> growth-related effects such as yeast cell ageing. In addition,<br />
nutritional effects including starvation, effect of oxygen, CO 2 <strong>and</strong> ethanol. These effects<br />
in combination have the potential to influence yeast physiology in ways that affect<br />
subsequent fermentation performance. Variable yeast physiology is not detected by<br />
viability tests.<br />
It is argued that to assess the cumulative effects of these influences on yeast it is<br />
necessary to use methods which are more discriminating than simple differentiation between<br />
viable <strong>and</strong> non-viable. The results of these tests may be used as the basis of a simple decision<br />
to use or discard yeast. Preferably they provide a result that is predictive of fermentation<br />
performance <strong>and</strong> they should identify appropriate values for parameters such as pitching rate<br />
<strong>and</strong> wort dissolved oxygen concentration that will provide optimum fermentation<br />
performance <strong>and</strong> consistent beer analysis. Several types of vitality test have been suggested<br />
which assess different aspects of yeast composition <strong>and</strong> biochemical function.<br />
The most relevant measures of yeast condition are those that directly reflect the<br />
changes that occur during fermentation. It is possible to perform laboratory fermentations<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
to predict performance at production scale, however, this is too time consuming to be of<br />
value. Rapid tests, which have been suggested, include measures of the rates of uptake of<br />
oxygen, uptake of glucose, evolution of CO2, ethanol formation <strong>and</strong> exothermy.<br />
Apparatus has been described that is capable of measuring some of these parameters. For<br />
example, the Bri vitality apparatus consists of an attemperated incubation chamber fitted<br />
with an integral dissolved oxygen probe (Kennedy, 1989). Providing a known<br />
concentration of yeast is placed within the chamber it is possible to measure specific<br />
rates of oxygen uptake. These are related to yeast sterol content <strong>and</strong> by implication<br />
provide a measure of the optimum oxygen requirement for efficient fermentation<br />
performance(Boulton<strong>and</strong>Quain,1987).Similarapparatusformeasuringspecificratesof<br />
CO2 evolution has been described (Muck <strong>and</strong> Narziss, 1988).<br />
The acidification power test measures the ability of yeast to acidify the medium both<br />
spontaneously (AP1) <strong>and</strong> in response to the addition of glucose (AP2) (Kara et al., 1988).<br />
Acidification occurs in response to proton extrusion <strong>and</strong> reflects the activity of plasma<br />
membrane H + ATPase. Maintenance of transmembrane potential requires metabolic<br />
activity, therefore the magnitude of AP1 reflects the availability of strorage<br />
carbohydrates. AP2 is fuelled by exogenous glucose <strong>and</strong> is indicative of glycolytic flux.<br />
The results of the acidification test as applied to pitching yeast correlate with subsequent<br />
fermentation performance (Fernadez et al., 1991; Mathieu et al., 1991; Siddique <strong>and</strong><br />
Smart, 2000).<br />
The acidification power test has been subject to several modifications. In one (Patino<br />
et al., 1993) maltose was substituted with glucose based on the fact that this is the major<br />
sugar component of wort. These authors suggested that the changes should be assessed<br />
eitherviathemeasureofconductance,orascumulatativeacidifcationpower ±thesum of<br />
changes in proton concentration over the course of the test. In another approach<br />
(Isenrentant et al., 1996) the amount of sodiumhydroxide required to maintain aconstant<br />
pH was determined. This was claimed to avoid the limitations of the st<strong>and</strong>ard method<br />
when using very high-vitality yeast.<br />
Several cell components, for example glycogen, trehalose <strong>and</strong> sterol, vary in<br />
concentration in response to changes in physiological condition. The concentrations of<br />
many of these cell components are known to be influential on fermentation performance.<br />
Variation from the norm in the concentrations of some of these cellular components is<br />
indicative of inappropriate yeast management.<br />
Metabolism is driven by the energy released by hydrolysis of ATP. The energy status of<br />
the cell is defined as the adenylate energy charge, which relates the relative concentrations<br />
of AMP, ADP <strong>and</strong> ATP (Chapman <strong>and</strong> Atkinson, 1977). The concentrations of these<br />
metabolites can be measured using bioluminescence (Section 17.3.1). The approach can be<br />
exp<strong>and</strong>ed to include the concentration of inorganic phosphate by nuclear magnetic<br />
resonance. Thereby, phosphorylation potential can be determined. Intracellular concentrations<br />
of glycogen <strong>and</strong> sterol are related to oxygen requirements in fermentation (Boulton <strong>and</strong><br />
Quain, 1987). Low concentrations of glycogen in pitching yeast are indicative of prolonged<br />
storage. Coupled with high sterol concentration, low glycogen is indicative of exposure to<br />
oxygen. Both metabolites can be determined by simple rapid tests. For example, glycogen<br />
by staining with iodine (Quain, 1981) <strong>and</strong> sterol by a spectrophotometric procedure using<br />
the polyene antibiotic, filipin (Rowe et al., 1991). Of these, determination of glycogen<br />
appears to be the most informative, particularly as an indicator of stressed yeast. The latter<br />
condition is associated with the presence of elevated concentrations of trehalose (12.5.7).<br />
Trehalose concentration can be rapidly determined by infra-red reflectance spectroscopy<br />
(Moonsamy et al., 1996).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Many aspects of cellular composition <strong>and</strong> physiological state are amenable to<br />
investigation using staining techniques. In particular, using biological fluorescent stains<br />
in conjunction with flow cytometry (Edwards et al., 1996; Lloyd <strong>and</strong> Dinsdale, 2000).<br />
Choice of an appropriate stain allows measurement of intracellular pH, glycogen<br />
concentration, geneological age, ploidy, membrane competence, budding index <strong>and</strong><br />
phase in cell cycle. The cost of flow cytometers continues to ensure that they are largely<br />
confined to research laboratories. Nevertheless, it is a very powerful technique <strong>and</strong> no<br />
doubt widespread adoption in quality assurance laboratories will be accompanied by a<br />
reduction in price.<br />
13.13 References<br />
AIRES BARROS, M. R., BARROS, M. R., CABRAL, J. M. S. <strong>and</strong> NOVAIS, J. M. (1987) Biotechnol. Bioeng., 24,<br />
1097.<br />
ASHURST, K. (1990) Brew. Dist. Internat., 21, 28.<br />
AUGUSTYN, O. P. H. <strong>and</strong> KOCK, J. F. L. (1989) J. Microbiol. Meth., 10, 9.<br />
BAMFORTH, C. W., BOULTON, C. A., CLARKSON, S. P. <strong>and</strong> LARGE, P. J. (1988) Proc. 20th Conv., Institute of<br />
<strong>Brewing</strong> (Australia <strong>and</strong> N. Zeal<strong>and</strong> Section), Brisbane, 209±218.<br />
BARFORD, J. P. (1987) `The Technology of Aerobic Yeast Growth'. In Yeast Technology, D. R. Berry, I.<br />
Russell <strong>and</strong> G. G. Stewart, eds, pp. 200±230, Allen & Unwin, Hemel Hempstead, UK.<br />
BARKER, M. G. <strong>and</strong> SMART, K. A. (1996) J. Amer. Soc. Brew. Chem., 54, 121.<br />
BOUIX, M. <strong>and</strong> LEVEAU, J.-Y. (2001) J. Inst Brew., 107, 217.<br />
BOULTON, C. A. <strong>and</strong> CLUTTERBUCK, V. J. (1993) Proc. 24th Cong. Eur. Brew. Conv, Oslo, 509.<br />
BOULTON, C. A. <strong>and</strong> QUAIN, D. E. (1987) Proc. 21st Cong. Eur. Brew. Conv., Madrid, 401.<br />
BOULTON, C. A. <strong>and</strong> QUAIN, D. E. (1999) Proc. 27th Cong. Eur. Brew. Conv., Cannes, 647.<br />
BOULTON, C. A. <strong>and</strong> QUAIN, D. E. (2001) <strong>Brewing</strong> Yeast <strong>and</strong> Fermentation, Blackwell Science, Oxford.<br />
BOULTON, C. A., MARYAN, P. S., LOVERIDGE, D. <strong>and</strong> KELL, D. B. (1989). Proc. 22nd Cong. Eur. Brew.<br />
Conv., Zurich, 653.<br />
BOULTON, C. A., BOX, W. G., CARVELL, J. <strong>and</strong> TURNER, K. (2001) Proc. 28th Cong. Eur. Brew. Cong.,<br />
Budapest, 722.<br />
BRANDL, J. (1996) Brauwelt Internat., 14, 32.<br />
BREITENBUCHER, K. <strong>and</strong> MISTLER M. (1995) Proc. EBC Monograph, XXIV, Espoo, 77.<br />
BURNS, J. A. (1941) J. Inst. Brew., 47, 10.<br />
CAHILL, G. <strong>and</strong> MURRAY, D. M. (2000) J. Amer. Soc. Brew., Chem., 58, 14.<br />
CASEY, G. P. (1996) MBAA Tech. Quart, 33, 1.<br />
CHAPMAN, A. G. <strong>and</strong> ATKINSON, D. E. (1977) Adv. Microb. Physiol., 15, 253.<br />
CHILVER, M. J., HARRISON, J. <strong>and</strong> WEBB, T. J. B. (1978) J. Amer. Soc. Brew. Chem., 36, 13.<br />
DEANS, K., PINDER, A, CATLEY, B. J. <strong>and</strong> HODGSON, J. A. (1997) Proc. 26th Cong. Eur. Brew. Conv.,<br />
Maastricht, 469.<br />
DE BARROS LOPES M., SODEN, A., MARTENS, A. L., HENSCHKE, P. A. <strong>and</strong> LANGRIDGE, P. (1998) Int. J.<br />
Systematic Bacteriol., 48, 279.<br />
DEBOURG, A. <strong>and</strong> VAN NEDERVELDE, L. (1999) Proc. 27th Cong. Eur. Brew. Conv., Cannes, 751.<br />
DEBOURG, A., LAURENT, M., GOOSENS, E., BORREMANS, F., VAN DER WINKEL, L. <strong>and</strong> MASSCHELEIN, C. A.<br />
(1994) J. Amer. Soc. Brew. Chem., 52, 100.<br />
DEERE, D., SHEN, J., VESEY, G., BELL, P., BISSINGER, P. <strong>and</strong> VEAL, D. (1998) Yeast, 14, 147.<br />
DINSDALE, M. G. <strong>and</strong> LLOYD, D. (1995) J. Inst. Brew., 101, 453.<br />
DINSDALE, M. G., LLOYD, D., MCINTYRE, P. <strong>and</strong> JARVIS, R. (1999) Yeast, 15, 285.<br />
EBC ANALYTICA MICROBIOLOGICA (1992) Vol. II, Method 3.1.1.2.<br />
EDWARDS, C., PORTER, J. <strong>and</strong> WEST, M. (1996) Ferment, 9, 288.<br />
ERNANDES, J. R., WILLIAMS, J. W., RUSSELL, I. <strong>and</strong> STEWART, G. G. (1993) J. Amer. Soc. Brew. Chem., 51,<br />
16.<br />
FELS, S., RECKELBUS, B. <strong>and</strong> GOSSELIN, Y. (1999) Proc. 7th IOB Conv. (Africa Section) Nairobi, 147.<br />
FERNANDEZ, S. S., GONZALEZ, G. <strong>and</strong> SIERRA, J. A. (1991) MBAA Tech. Quart., 28, 89.<br />
FUGE, E. K., BRAUN, E. I. <strong>and</strong> WERNER-WASHBURNE, M. (1994) J. Bacteriol., 176, 5802.<br />
GALAZZO, J. L. <strong>and</strong> BAILEY, J. E. (1989) Biotechnol. Bioeng., 33, 1283.<br />
GALAZZO, J. L. <strong>and</strong> BAILEY, J. E. (1990) Enz. Microb. Technol., 12, 162.<br />
GEIGER, E. (1993) Brauwelt Internat, 11, 430.<br />
GERVAIS, P., MARCHAL, P. A. <strong>and</strong> MOLIN, P. (1992). Biochim., Biophys., Acta, 40, 1435.<br />
GILLILAND, R. B. (1951) Proc. 3rd Cong. Eur. Brew. Conv., Brighton, 35.<br />
GODIA, F., CASAS, C. <strong>and</strong> SOLA, C. (1987) Proc. Biochem., April, 43.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
GOODACRE, R. (1994) Microbiol. Europe, 2, 16.<br />
GULDFELDT, L. U. <strong>and</strong> PIPER, J. U. (1999). MBAA Tech. Quart., 36, 1.<br />
HALL, J. F. (1954) J. Inst. Brew., 60, 482.<br />
HARRIS, C. M., TODD, R. W., BUNGARD, S. J., LOVITT, S. J., MORRIS, J. G. <strong>and</strong> KELL, D. B. (1987) Enz.<br />
Microbiol. Technol., 9, 181.<br />
HAYFLICK, L. (1965) Expt. Cell Res., 37, 614.<br />
HEGGART, H. M., MARGARITIS, A., PILKINGTON, H., STEWART, R. J., DOWHANICK, T. M. <strong>and</strong> RUSSELL, I.<br />
(1999). MBAA Tech. Quart., 36, 383.<br />
HELM, D., LABISCHINSKI, H., SCHALLEHN, G. <strong>and</strong> NAUMANN, D. (1991) J. Gen. Microbiol., 137, 69.<br />
HELMS, E., NOHR, B. <strong>and</strong> THORNE, R. S. W. (1953) Wallerst. Lab. Commun., 16, 115.<br />
HOLCBERG, I. B. <strong>and</strong> MARGALITH, P. (1981) Eur. J. Appl. Microbiol., Biotechnol., 13, 133.<br />
HOUGH, J. S. (1957) J. Inst. Brew., 63, 483.<br />
HOUGH, J. S., BRIGGS, D. E., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1982) Malting <strong>and</strong> <strong>Brewing</strong> Science, Vol. 2.,<br />
Hopped Wort <strong>and</strong> Beer, (2nd edn), Chapman <strong>and</strong> Hall, London <strong>and</strong> New York.<br />
IOB METHODS OF ANALYSIS (1997) Vol. 2, Microbiology, Inst. Brew., Clarges St., London.<br />
ISENRENTANT, D., GREENANS, W. <strong>and</strong> VERACHTERT, H. (1996) J. Am. Soc. Brew. Chem., 54, 110.<br />
JAZWINSKI, S. M. (1999) Expt. Gerontology, 34, 1.<br />
JONES, R. P. (1987) Proc. Biochem., 22, 130.<br />
KAMATH, R. S. <strong>and</strong> BUNGAY, H. R. (1988) J. Gen. Microbiol., 134, 3061.<br />
KARA, B. V., SIMPSON, W. J. <strong>and</strong> HAMMOND, J. R. M. (1988) J. Inst. Brew., 94, 153.<br />
KENNEDY, R. (1989) Brew. Guard., Sept, 57.<br />
KIRSOP, B. E. (1991) `Maintenance of yeasts'. In Maintenance of Microorganisms <strong>and</strong> Cultured Cells, A<br />
Manual of Good Practice, 2nd edn, B. E. Kirsop <strong>and</strong> A. Doyle, eds, pp. 161±182, Academic Press,<br />
London.<br />
KOSHCHEYENKO, K. A., TURKINA, M. V. <strong>and</strong> SKRYABIN, G. K. (1983) Enz. Microb. Technol., 5, 14.<br />
KRIKILION, P. H., ANDRIES, M., GOFFIN, O, VAN BEURAN, P. C. <strong>and</strong> MASSCHELEIN, C. A. (1995). Proc. 25th<br />
Cong. EBC, Brussels, 419.<br />
KURLANZKA, A., RYTKA, J., ROZALSKA, B. <strong>and</strong> WYSOCKA, M. (1999) Yeast, 15, 23.<br />
LENTINI, A. A. (1993) Ferment, 6, 321.<br />
LLOYD, D. (1993) Flow Cytometry in Microbiology, D. Lloyd ed., Springer Verlag, Berlin.<br />
LLOYD, D. <strong>and</strong> DINSDALE, G. (2000) `From bright field to fluorescent <strong>and</strong> confocal microscopy'. In<br />
<strong>Brewing</strong> Yeast Fermentation Performance, K. A. Smart, ed., pp. 3±9, Blackwell Science, Oxford.<br />
LLOYD, D. <strong>and</strong> HAYES, A. J. (1995) FEMS Microbiol. Lett., 133, 1.<br />
MADEO, F., FROHLICH, E. <strong>and</strong> LIGER, M. (1999) J. Cell Biol., 145, 757.<br />
MASSCHELEIN, C. A., BORREMANS, E. <strong>and</strong> VAN DE WINKEL, L. (1994) Proc. 23rd IOB Conv. (Asia Pacific<br />
Section), Sydney, 109.<br />
MATHIEU, C., VAN DEN BERGH, L. <strong>and</strong> ISERENTANT, D. (1991) Proc. 23rd EBC Conv., Lisbon, 273.<br />
MCCAIG, R. (1990) J. Amer. Soc. Brew., Chem., 48, 22.<br />
MCMURROUGH, I. (1995) Proc. EBC Monograph, XXIV, Espoo, 2.<br />
MIEDENER, H. (1978) Proc. EBC Conv, Symp. Monograph V, Zouterwoude, 110.<br />
MOONSAMY, N., MOCHOBA, F., O'CONNOR-COX, E. S. C. <strong>and</strong> AXCELL, B. C. (1996) J. Inst. Brew., 101, 203.<br />
MORRIS, G. J., COULSON, G. E. <strong>and</strong> CLARKE, K. J. (1988) Cryobiology, 25, 471.<br />
MOZES, N., MARCHAL, F. <strong>and</strong> HERMESSE, M. P. (1987) Biotechnol. Bioeng., 30, 439.<br />
MUCK, E. <strong>and</strong> NARZISS, L. (1988) Brauwelt Internat., 1, 61.<br />
NAUDTS, G., AERTS, G. <strong>and</strong> ISERENTANT, D. (1997) Proc. 26th Cong. Eur. Brew. Conv, Maastricht, 377.<br />
NIELSEN, H., HOYBE-HANSEN, I., IBACK, D. <strong>and</strong> KRISTENSEN, B. J. (1986) Brygmesteren, 2, 7.<br />
NIELSEN, H., HOYBE-HANSEN, I., IBACK, D., KRISTENSEN, B. J. <strong>and</strong> SYNNESVEDT, K. (1987) MBAA Tech.<br />
Quart., 24, 90.<br />
OH, K.-B. <strong>and</strong> MATSUOKA, H. (2002) Int. J. Food. Micrbiol., 76, 47.<br />
OWADES, J. L. (1981) MBAA Tech. Quart., 18, 163.<br />
PAJUNEN, E. (1995) Proc. EBC Monograph, XXIV, Espoo, 24.<br />
PAJUNEN, E. <strong>and</strong> GRONQVIST, A. (1994) Proc. 23rd IOB Conv. (Australia & New Zeal<strong>and</strong> Section),<br />
Sydney, 101.<br />
PARKINNEN, E., OURA, E. <strong>and</strong> SOUMALAINEN, H. (1976). J. Inst. Brew., 82, 283.<br />
PATINO, H., EDELSEN, C. <strong>and</strong> MILLER, J. (1993) J. Amer. Soc. Brew. Chem., 51, 128.<br />
PIRT, S. J. (1975) Principles of Microbe <strong>and</strong> Cell Cultivation, Blackwell Scientific Publications, Oxford.<br />
POSADA, J. (1978) Proc. EBC Conv, Symp. Monograph V, Zouterwoude, 207.<br />
POWELL, C. A., VAN ZANDYCKE, S. M., QUAIN, D. E. <strong>and</strong> SMART, K. A. (1999) Microbiology, 146, 1023.<br />
QUAIN, D. E. (1981) J. Inst. Brew., 87, 289.<br />
QUAIN, D. E. (1995) Proc. 25th Cong. EBC, Brussels, 309.<br />
QUAIN, D. E. <strong>and</strong> TUBB, R. S. (1982) MBAA Tech. Quart., 19, 29.<br />
QUAIN, D. E., BOX, W. G. <strong>and</strong> WALTON, E. F. 1985) Lab. Practice, 34, 84.<br />
QUAIN, D. E., POWELL, C., HAMILTON, A., RUDDLESDEN, D. <strong>and</strong> BOX, W. (2001) Proc. 28th Cong. Eur.<br />
Brew. Conv., Budapest, 388.<br />
RAYNAL, L., BARNWELL, P. <strong>and</strong> GERVAIS, P. (1994) J. Biotechnol., 36, 121.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
REISS, S. (1986) MBAA Tech. Quart., 23, 32.<br />
ROWE, S .M., SIMPSON, W. J. <strong>and</strong> HAMMOND, J. R. M. (1991) Lett. Appl. Microbiol., 13, 182.<br />
RUSSELL, I. <strong>and</strong> STEWART, G. G. (1981) J. Amer. Soc. Brew. Chem., 39, 19.<br />
RYDER, D. S., BOWER, P. A. <strong>and</strong> BROMBERG, S. K. (1995) Proc. EBC Monograph, XXIV, Espoo, 175.<br />
SCHMIDT, H.-J. (1995) Brauwelt Internat., 13, 130.<br />
SCHOFIELD, M. A., ROWE, S. M., HAMMOND, J. R. M., MOLZAHN, S. W. <strong>and</strong> QUAIN, D. E. (1995) J. Inst. Brew.,<br />
101, 75.<br />
SEDDON, A. W. (1975) MBAA Tech. Quart., 3, 130.<br />
SIDDIQUE, R. <strong>and</strong> SMART, K. (2000) `Predicting fermentation performance using proton efflux', In,<br />
<strong>Brewing</strong> Yeast Fermentation Performance, K. A. Smart, ed., pp. 46±54, Blackwell Science, Oxford.<br />
SIGSGAARD, P. <strong>and</strong> RASMUSSEN, J. N. (1985) J. Amer. Soc. Brew. Chem., 43, 104.<br />
SKANDS, B. (1997) Proc. 26th Cong. Eur. Brew. Conv., Maastricht, 413.<br />
SMART, K. A. <strong>and</strong> WHISKER, S. (1996) J. Am. Soc. Brew. Chem., 54, 41.<br />
SMART, K. A., CHAMBERS, K. M., LAMBERT, I. <strong>and</strong> JENKINS, C. (1999) J. Am. Soc. Brew. Chem., 57, 180.<br />
STEWART, G. G., ZHENG, X. <strong>and</strong> RUSSELL, I. (1995) Proc. 25th Cong. EBC, Brussels, 403.<br />
STRATTON, M. K., CAMPBELL, S. J. <strong>and</strong> BANKS, D. J. (1994) Proc. 23rd IOB Conv. (Asia Pacific Section),<br />
Sydney, 96.<br />
THIBAULT, J., LEDAY, D. <strong>and</strong> COTE, F. (1987) Biotech. Bioeng., 30, 74.<br />
TIMMINS, E. M., QUAIN, D. E. <strong>and</strong> GOODACRE, R. (1998) Yeast, 14, 885.<br />
TUBB, R. S. <strong>and</strong> LILJESTROÈ M, P. L. (1986) J. Inst. Brew., 92, 588.<br />
VAN DE WINKEL, L., VAN BEVERAN, P. C. <strong>and</strong> MASSCHELEIN, C. A. (1991) Proc. 23rd Cong. EBC, Lisbon,<br />
577.<br />
VAN VUUREN, H .J. J. <strong>and</strong> VAN DER MEER, L. (1987) Amer. J. Enol. Viticult., 38, 49.<br />
WALSH, R. M. <strong>and</strong> MARTIN, P. A. (1977) J. Inst. Brew., 83, 169.<br />
WERNER-WASHBURNE, M., BRAUN, E. I., CRAWFORD, M. E. <strong>and</strong> PECK, V. M. (1996) Molec. Microbiol., 1159.<br />
WESTNER, H. (1999) Brew. Dist. Internat., Sept. 24.<br />
WILLIAMS, R. P. <strong>and</strong> RAMSDEN, R. (1963) Continuous fermentation process <strong>and</strong> apparatus for beer<br />
production. British Patent 926847.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
14<br />
Fermentation technologies<br />
14.1 Introduction<br />
The object of brewery fermentation is to utilize the ability of yeast cells to convert sugar<br />
into ethanol <strong>and</strong> carbon dioxide as the major products of metabolism. The yeast also<br />
produces aseries of minor metabolites such as esters, higher alcohols <strong>and</strong> acids that<br />
contributepositively toflavour.Itisthesemyriadsofminor components thatcharacterize<br />
abeer br<strong>and</strong> <strong>and</strong> make it identifiable to adrinker. The chosen yeast must also control the<br />
elimination of undesirable flavour components arising from the raw materials or from<br />
fermentation. Much of this flavour improvement occurs in maturation (Chapter 15).<br />
Brewery fermentations are discussed in Chapter 12. In this Chapter the technology of<br />
fermentation is described with the objective of performing the process consistently to<br />
yield the highest quality beer at lowest cost. Low costs must be demonstrated as low<br />
capital cost to keep the depreciation element of brewery fixed costs as low as is<br />
achievable <strong>and</strong> as low maintenance cost. Minimum beer losses must also be achieved so<br />
that the maximum amount of beer is derived from the raw materials used. In this way<br />
brewery variable cost per unit volume of beer produced is lowered. To achieve these<br />
objectives requires careful selection of the yeast <strong>and</strong> the appropriate wort composition<br />
(Chapter 12). It also requires efficient operation of properly designed fermentation<br />
equipment. In this respect top <strong>and</strong> bottom fermenting yeasts must be considered.<br />
The nature of fermentation was not clearly established until the latter part of the<br />
nineteenth century <strong>and</strong> so the design of vessels to effectively carry out fermentation has a<br />
history of about 150 years. This design is fundamentally influenced by the mode of<br />
separation of the yeast, i.e., top or bottom fermenting yeasts. Cool, bottom fermentation<br />
was almost entirely restricted to Bavaria until 1840 when the rest of the world was using<br />
top fermentation (Christian, 1959). Some bottom fermenting yeast was smuggled to<br />
Czechoslovakia in 1842 <strong>and</strong> so Pilsen was established as a major brewing centre. About<br />
three years later a Danish brewer, J. C. Jacobsen, took bottom fermenting yeast from<br />
Munich to Copenhagen <strong>and</strong> so the Danish city also became established as important for<br />
brewing. Of great significance at about the same time was the introduction of a bottom<br />
fermenting yeast to Pennsylvania in the USA. Its use spread rapidly through the actions of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
immigrant German brewmasters such as Frederick Miller, Bernhard Stroh, Eberhard<br />
Anheuser <strong>and</strong> Adolph Coors. The use of bottom fermentation was thus spreading<br />
throughout Europe <strong>and</strong> North America at the same time as Pasteur began his<br />
microbiological research resulting in the development of atheory of fermentation, culture<br />
techniques for micro-organisms, <strong>and</strong> the principles of sterilization (Pasteur, 1860, 1876).<br />
Progress in controlling brewery fermentation could now be made as improvements in the<br />
microbiologicalaspectsoffermentationcouldfacilitateengineeringdevelopments.Ofcritical<br />
importanceinbottomfermentationwastemperaturecontrol<strong>and</strong>inparticularthemaintenance<br />
oflowtemperatures(0ëC,32ëF<strong>and</strong>less)forlongperiodsoftime.Thisoriginallyrequiredthe<br />
use of large quantities of ice until the compressor refrigerator appeared in breweries from<br />
about1873followingdevelopmentsinAustralia<strong>and</strong>Germany.Thenext100yearssawthe<br />
ab<strong>and</strong>onment of top fermentation throughout the world with the exception of the UK <strong>and</strong><br />
Irel<strong>and</strong><strong>and</strong>rapiddevelopmentsinthelateryearsofthetwentiethcenturytoincreasedbatch<br />
size fermentations in very large vessels (up to 6,000hl, 3,600imp. brl).<br />
There has recently been renewed interest in top fermented ale beers produced in small<br />
batches in `niche' breweries particularly in the USA. There has also been adesire to<br />
eliminate differences between top <strong>and</strong> bottom fermentation by seeking semi-continuous<br />
<strong>and</strong> continuous fermentation techniques, some of them employing immobilized yeast,<br />
when the traditional <strong>practice</strong>s of top <strong>and</strong> bottom separation do not take place.<br />
Therearemarkeddifferencesinthebatchsizesoffermentationsystemsthroughoutthe<br />
world. This reflects the market in which the particular brewer operates. International<br />
brewers producing global br<strong>and</strong>s will have plants with capacities of at least 10m. hl per<br />
annum(6m.imp.brl)<strong>and</strong>sometimesconsiderablygreaterthanthisintheUSA<strong>and</strong>Japan.<br />
Regional brewers in many countries brew successfully in plants of annual production in<br />
the range 0.1 to 1.0m. hl (60,000 to 600,000imp. brl) <strong>and</strong> craft or niche brewers would<br />
operate at levels of 1,000 to 10,000hl per annum. Fermentation processes are therefore<br />
successfully carried out in batch sizes ranging from 10 to 6,000hl. It follows that awide<br />
range of vessel types are used.<br />
Techniques of fabrication are available to make fermenting vessels of all sizes.<br />
Practical sizes are limited by economic factors. It is difficult to transport by road, rail or<br />
barge, vessels much bigger than 2,000hl capacity (1,200imp. brl) <strong>and</strong> vessels of greater<br />
volume must be built on site at greater cost. Companies must decide on the optimum size<br />
for their own operations. Fermentation technology, therefore, embraces astudy of:<br />
· fermenters for bottom <strong>and</strong> top fermentations <strong>and</strong> consequent yeast separation<br />
· fermentersforcontinuous<strong>and</strong>semi-continuousoperationrequiringnoyeastseparation<br />
· fermentation control systems.<br />
Some basic principles first need outlining so that vessel design <strong>and</strong> operation is set in the<br />
context of the biochemical requirements of successful brewery fermentation (see also<br />
Chapters 12, 13).<br />
14.2 Basic principles of fermentation technology<br />
14.2.1 Fermentability of wort<br />
The main objective is to ferment wort to the desired gravity; this is often called the<br />
required degree of attenuation. The proportion of the wort dissolved solids (extract)<br />
which can be fermented is called the percentage fermentability of the wort:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Original gravity Final gravity<br />
Fermentability…%† ˆ<br />
Original gravity<br />
The original gravity can be expressed in ëP (Plato), which measures the concentration in<br />
weight/weight terms as grammes of solids per 100 grammes of wort, or in ëSacch, which<br />
relates thespecific gravityoftheworttothatofwater takenas1,000.Finalgravitymeans<br />
thegravityofthewortwhenitisfullyfermentedsuchthataddingmoreyeastorleavingit<br />
longer will lead to no further fall in gravity. This lowest gravity is often called the<br />
attenuation limit gravity <strong>and</strong> when it is reached the beer is said to be fully attenuated.<br />
Therefore if wort of original gravity 48ëSacch (12ëP) is fermented to a gravity of<br />
12ëSacch (3ëP) the percentage fermentability is:<br />
‰48 12=48Š 100 ˆ75% or ‰12 3=12Š 100 ˆ75%<br />
The alcohol formed in the fermentation has alower density than water <strong>and</strong> so it<br />
decreases the final gravity. The final gravity, therefore, does not show the amount of<br />
extract left in the fermented wort. The attenuation limit gravity referred to above is<br />
therefore called the apparent attenuation limit <strong>and</strong> what is calculated by the equation is<br />
the apparent attenuation of the wort. To measure the real attenuation the alcohol must be<br />
removed, e.g., by distillation before determining the gravity. The real attenuation is<br />
approximately 80% of the apparent attenuation. The true factor published by Balling in<br />
1880 was 0.81, in modern <strong>practice</strong> the real attenuation can be obtained from the apparent<br />
attenuation by the use of tables.<br />
To determine the degree of attenuation achievable from agiven wort the attenuation<br />
limit is usually measured in the laboratory. Yeast, pre-washed with wort, is mixed with<br />
filtered wort, which is fermented at 25ëC. Aspecific gravity reading is taken after two<br />
days <strong>and</strong> then repeated twice aday until the gravity does not change. This gravity is the<br />
apparent attenuation limit. It is the highest apparent degree of attenuation that can be<br />
achieved by fermentation of all fermentable material in the extract of the wort. It is, of<br />
course, governed by the raw materials chosen to make the wort <strong>and</strong> the extent of enzyme<br />
activity in the brewhouse.<br />
In the brewery fermentation the apparent attenuation limit is not usually reached.<br />
Some brewers set aspecific attenuation limit for aparticular beer <strong>and</strong> attempt to reach<br />
this value in apreset time against apreset fermentation temperature regime. If there is a<br />
large difference between the final attenuation achieved <strong>and</strong> the apparent attenuation limit<br />
as determined in the laboratory then there is fermentable extract present in the beer <strong>and</strong><br />
this represents a risk of supporting infection by yeasts <strong>and</strong> bacteria. This would<br />
subsequently cause the beer to go hazy <strong>and</strong> would create off-flavours. Generally, in<br />
modern large batch lager fermentations, the objective is to try to ferment the beer to the<br />
limit. This is not the object in the production of cask ale in the UK when residual<br />
carbohydrate is required in the beer to allow secondary fermentation <strong>and</strong> cask<br />
conditioning to take place (Chapter 21).<br />
14.2.2 Time course of fermentation<br />
Bydeterminingthespecificgravityofthewortattimeintervalsonecanfollowthecourse<br />
of fermentation. Typical fermentation profiles for ale <strong>and</strong> lager fermentations are shown<br />
in Fig. 14.1, which also includes typical temperature regimes although these vary<br />
between breweries. The decline in the specific gravity is matched by the growth of the<br />
yeast as sugars are metabolized <strong>and</strong> ethanol produced. The pH value of the wort falls as<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
100
Fig. 14.1 Time course of fermentation for ale (a) <strong>and</strong> lager (b) beers. fa, level of fusel alcohols<br />
( g/l); e, level of esters ( g/l) t, temperature ëC. (Hough et al., 1982).<br />
ammonium ions <strong>and</strong> amino acids are taken from the wort by the yeast <strong>and</strong> organic acids<br />
are secreted (Chapter 12). Major flavour compounds, the esters <strong>and</strong> higher alcohols are<br />
released into the wort as the yeast grows <strong>and</strong> so increase in concentration as fermentation<br />
proceeds.<br />
The concentration of these flavour compounds is critical for the consistency of the<br />
beer br<strong>and</strong> <strong>and</strong> the flavour profile of the beer (Chapter 20). Brewers need to be aware of<br />
the decline in concentration of flavour compounds, which can occur towards the end of<br />
fermentation, <strong>and</strong> in maturation (Chapter 15). This can occur as volatiles are carried out<br />
with evolving carbon dioxide (`gas purging') or are re-absorbed by the yeast. Marked<br />
differences can be seen in the time course profiles of ale <strong>and</strong> lager fermentations <strong>and</strong><br />
these differences are reflected in the flavour profiles of the resulting beers. It follows that<br />
the vessels for fermentation must be equipped to regulate these differences consistently<br />
so that consistent products can be brewed.<br />
14.2.3 Heat output in fermentation<br />
Fermentation is an exothermic process. The metabolism of wort sugars in brewery<br />
fermentation can be represented in general by the summation of the biochemical pathway<br />
discussed in Chapter 12:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Glucose‡2Pi ‡2ADP !2Ethanol‡2CO2 ‡2ATP<br />
In this version of the equation energy is fixed <strong>and</strong> stored by the yeast cell as ATP. The<br />
free energy of reaction ( G) can be calculated for each stage of the process (Mahler <strong>and</strong><br />
Cordes, 1969) <strong>and</strong> anet figure of 157 kJ/mole can be derived. However, the ATP<br />
generated is used in further reactions in the cell:<br />
2ATP !2ADP +2Pi Gˆ2 31 ˆ62kJ/mole<br />
Therefore the overall heat of production can be assessed as 157‡62 ˆ219kJ/mole of<br />
glucose fermented. If temperature is to be controlled during fermentation to ensure a<br />
consistent performance then this heat must be removed. The heat produced during a<br />
typical fermentation can thus be assessed from this figure (Anderson et al., 2000).<br />
Consider a12ëP wort; this will contain 12.6kg of extract/hl of wort of which, say,<br />
75% will be fermentable. The molecular weight of glucose is 180 <strong>and</strong> so 219kJ/mole of<br />
glucose is equivalent to …219=180† 1000kJ/kg of glucose, i.e., 1217kJ. Therefore<br />
fermentation of 1000hl of wort would yield in total:<br />
1000 …12:6 75=100† 1217 ˆ11:5GJ<br />
The heat is not uniformly given out but will reach apeak at maximum fermentation rate.<br />
This peak has been variously estimated but avalue with practical worth, is 0.22kg<br />
extract/hl/h (Fricker, 1978). The maximum cooling load for a1000hl fermenter would<br />
therefore be:<br />
1000 0:22 1217 ˆ0:26 GJ/h<br />
This heat must be removed from the fermenter if the temperature of fermentation is to be<br />
controlled. The amount of cooling required depends on the temperature range to be<br />
maintained. Frequently in the production of ales with rapid top fermentation the<br />
temperature is allowed to rise unrestricted over the first 48 to 60 hours of fermentation.<br />
Temperatures of over 22ëC (72ëF) can be reached. Cooling is usually then applied to<br />
lower the temperature to around 15ëC (59ëF) for abrief maturation (Chapter 15) before<br />
proceeding to conditioning or to cask racking. In a lager fermentation temperature of the<br />
fermenting beer is usually controlled so that it rises to no more than 13 ëC (55 ëF) after<br />
pitching at around 8 ëC (46 ëF) <strong>and</strong> is then lowered by progressive cooling to about 5 ëC<br />
(41 ëF). The beer will then proceed to the maturation <strong>and</strong> conditioning stage. An equation<br />
has been derived to calculate the amount of cooling required which can be applied to any<br />
fermentation situation (Anderson, et al., 2000):<br />
Q ˆ M Cp t<br />
where Q is the heat in kJ, M is the mass of the beer in kg, Cp is the specific heat in kJ/kg/ëC,<br />
<strong>and</strong> t is the required temperature change of the beer. Therefore if a fermenter contains<br />
1000 hl of beer with a specific heat of 4.05 <strong>and</strong> it is required to lower temperature from<br />
13 ëC to 5 ëC, the heat removed will approximately be (assuming gravity to be 1.00):<br />
1000 100 (13 5) 4.05 = 3.24 GJ<br />
There is frequently debate between brewers <strong>and</strong> equipment suppliers about the rate of<br />
cooling required to remove this quantity of heat. To control temperature during<br />
fermentation a rate of temperature reduction of 1 ëC/hour is often used. In the above<br />
example this would equate to:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1000 100 3.24 =324 MJ/h<br />
Some brewers require arate greater than this to `crash cool' afermenter to precipitate<br />
yeast <strong>and</strong> achieve aconditioning temperature of
Conical nozzle<br />
with sightglass<br />
Jacket outlet<br />
Conical jacket<br />
outlet<br />
Thermometer<br />
CO2 injection<br />
cock<br />
Outlet<br />
cock<br />
Yeast cock<br />
with sightglass<br />
Pipe for CO2<br />
entry <strong>and</strong><br />
pressure cleaning<br />
Jacket inlet<br />
Vessel cleaning<br />
<strong>and</strong> pressure<br />
delivery pipe<br />
CO2 washing<br />
lantern<br />
Conical jacket inlet<br />
Fig. 14.2 Cylindroconical fermentation vessel, (Hough et al., 1982).<br />
These vessels are usually 3±4 times taller than their diameter <strong>and</strong> work at pressures of<br />
1±1.5 bar above atmospheric. Tank diameters in the UK are usually 3.5 to 4.5m(11.5<br />
to 15ft.). Heterogeneous fermentations have been observed in tanks much greater than<br />
20m(66ft.) high <strong>and</strong> for this reason in recent developments the height of vessels has<br />
been kept to less than 15m(49ft.). This phenomenon has not been fully explained but<br />
special circumstances do apply to very large vessels (>2,500hl 1,500imp. brl) in<br />
relation to temperature gradients <strong>and</strong> cooling (Section 14.3.3). In classic European<br />
lager production more squat vessels are used where the ratio of diameter to height is<br />
3:1, there is atendency for<br />
increased production of higher alcohols at the expense of esters. This may be caused<br />
by increased amino acid utilization caused by the increased beer circulation generated<br />
byrapidcarbondioxideproduction.Increasingthesizeofthevessel canlowercostper<br />
unit volume. Doubling the size of the vessel leads to acost increase of about 35%<br />
(Maule,1977).Ingeneralthegreaterthevolumetosurfacearearatiothelowertheunit<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
volume cost will be. Large vessels are sited in the open. These are subject to planning<br />
regulations in some countries.<br />
Cylindroconical vessels cannot be completely filled for primary fermentation. Alarge<br />
volume of foam is formed by the evolution of carbon dioxide <strong>and</strong> this could cause<br />
pressure release valves to block. The headspace volume of the tank should therefore be at<br />
least 25% of the pitched wort volume. From this discussion it is difficult to generalize<br />
about the optimum size <strong>and</strong> aspect ratio of fermenters. As ageneral rule filling the vessel<br />
shouldnot take longer than12 hours, irrespective ofthenumberofbrews beingruntothe<br />
vessel for fermentation. Continuous filling with fresh wort will lead to increased<br />
production of -acetolactate <strong>and</strong> result in enhanced maturation time for diacetyl removal<br />
(Chapter 15). It is useful in abrewery to have arange of sizes of vessels. If the brew<br />
length is 1000 hl (600 imp. brl) then vessels of 4,000±5,000 hl (2,400±3,000 imp. brl)<br />
capacity will be suitable but it will be advantageous to have vessels of 250 hl (150 imp.<br />
brl) size for propagation <strong>and</strong> some 500±1,000 hl (300±600 imp. brl) vessels for new<br />
product development <strong>and</strong> trial work.<br />
14.3.2 Construction of cylindroconical vessels<br />
Originally fermenting vessels for bottom fermentation were constructed of mild steel<br />
with a glass or epoxy resin lining. This lining had to be frequently inspected to ensure its<br />
integrity. Mild steel was also prone to rusting <strong>and</strong> modern vessels are almost always<br />
constructed of chrome-nickel stainless steels.<br />
Metals <strong>and</strong> design<br />
Generally the steels used are stainless <strong>and</strong> austenitic, i.e., containing carbon, which forms<br />
a carbide with the gamma form of iron, which is normally only stable at high<br />
temperatures but can be stabilized at normal temperatures by the inclusion of elements<br />
such as chromium, nickel, <strong>and</strong> molybdenum. These steels are often referred to in brewing<br />
by the general classification V2A <strong>and</strong> V4A, but these categories cover a series of<br />
different alloys. The class can be subdivided into AISI 304 (V2A group) <strong>and</strong> AISI 316<br />
(V4A group), which are the steels in most common use. Resistant properties of 316 steels<br />
are enhanced by the inclusion of molybdenum (Table 14.1). A further category of steels,<br />
designated 321 contain titanium. Normally 304 stainless is used. However V2A steels are<br />
not fully resistant to chloride ions or to pH values < 4.5. This is not usually a problem for<br />
fermenters but with liquors having high chloride contents 316 can be specified, but it is<br />
much more expensive than 304 steel.<br />
A very important factor is the surface smoothness of the steel that can be achieved in<br />
manufacture. It should be as smooth as is possible so that indentations cannot provide areas<br />
for potential microbial contamination. Cold rolling of stainless plate will yield a surface<br />
with a `roughness' (Ra) of 0.6 to 0.8 m. In some systems of work this type of finish is<br />
designated `2b'. Electro-polishing can lower Ra to 0.3 to 0.4 m but this will be at greater<br />
cost <strong>and</strong> is not always specified by brewers. Vessels are now almost always produced to<br />
Table 14.1 Composition of austenitic stainless steels (Barnes, 2001)<br />
Type Carbon (%) Chromium (%) Nickel (%) Molybdenum (%)<br />
304 0.03±0.06 17.5±19.0 8.0±12.0 ±<br />
316 0.03±0.07 16.5±18.5 10.0±14.0 2.25±3.0<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
st<strong>and</strong>ards of design that can be specified by the brewer at the time of making an enquiry to<br />
purchase. Choice is thus facilitated which can concentrate on price aspects. Generally three<br />
design st<strong>and</strong>ards are recognized in Europe <strong>and</strong> North America: AD MerkblaÈtter (Germany),<br />
ASME (American Society of Mechanical Engineers) VII div 1(USA), <strong>and</strong> BS 5500 (UK).<br />
Vessels operating at pressures >0.5 bar will normally be subject to national regulations<br />
relating to pressure vessels <strong>and</strong> will require inspections for insurance purposes.<br />
Cooling jackets<br />
The fermenter must be equipped with acooling system to remove heat generated during<br />
fermentation <strong>and</strong> to allow the control of temperature to the required profile. If the vessel<br />
is to be used for maturation as well (Section 15.2.5) then there is the requirement to cool<br />
to less than 0ëC (32ëF) <strong>and</strong> to hold this temperature for several days. Systems of direct<br />
<strong>and</strong> indirect cooling can be used. In direct cooling ammonia gas is used as the refrigerant<br />
<strong>and</strong> cooling is achieved by expansion <strong>and</strong> evaporation of the liquefied gas. In indirect<br />
cooling asecondary coolant (such as IMS, industrial methylated spirit), generated from a<br />
refrigeration plant is used. This coolant is circulated through the fermenting vessels <strong>and</strong><br />
thencereturnedtotheplant.Coolingjacketsareconstructedsothatgoodheatexchangeis<br />
possible <strong>and</strong> several designs are available (Fig. 15.2). The choice often depends on the<br />
expertise of the proposed manufacturer. Limpet coils are wound onto the vessel surface<br />
<strong>and</strong> seam welded, they have a relatively heavy construction. An alternative is the pressed<br />
corrugated `profile' plate (Barnes, 2001), which can be pre-formed <strong>and</strong> spot welded to the<br />
vessel shell. In some versions this is called a `dimple' jacket. There is the further<br />
alternative of a `quilted' panel, which is hydraulically inflated after spot welding to the<br />
vessel surface. This system has a low volume <strong>and</strong> is favoured with direct expansion<br />
cooling using ammonia. If properly specified <strong>and</strong> manufactured the design of the cooling<br />
jacket has little impact on fermentation performance. A fully equipped fermenter will<br />
probably have two cooling sections on the cylinder of the tank <strong>and</strong> a further section on<br />
the cone.<br />
Vessel fittings<br />
Vessels are filled <strong>and</strong> emptied from below, reducing oxygen ingress. Vessels are fitted<br />
with pipes for the addition of wort, the removal of yeast, <strong>and</strong> the removal of beer. There<br />
are also pressure relief <strong>and</strong> vacuum relief systems, which are fitted into a top-plate<br />
assembly. Cleaning-in-place (CIP) fluids must also be introduced <strong>and</strong> this is usually<br />
through the top plate although, for ease of access, the valve may be situated at the bottom<br />
of the tank.<br />
One of the main factors in the operation of these vessels is ensuring the integrity of the<br />
pipe-work systems so that yeast, wort, beer <strong>and</strong> cleaning fluids are, when required,<br />
h<strong>and</strong>led separately <strong>and</strong> not allowed to mix. This is frequently achieved by using a `dial-apipe'<br />
system where connections are made manually. The different lengths of the Ushaped<br />
connection pipes make wrong connections impossible. The pathways so made are<br />
opened <strong>and</strong> closed by manual or remotely operated butterfly valves.<br />
In large modern fermenting `tank farms' the vessels are often permanently connected<br />
by the pipe-work system. Valves are then either connected to every tank or collected into<br />
a `valve routeing block' to which every tank is connected <strong>and</strong> which is automatically <strong>and</strong><br />
remotely controlled. The problem in this system is valve leakage which can cause severe<br />
damage if, say, cleaning fluid leaks into the beer. The risk is reduced by the use of<br />
`double seat' valves that provide leakage indication. This assembly contains an upper <strong>and</strong><br />
lower valve controlled by springs <strong>and</strong> separated by a small space. Opening the valve is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
achieved by compressed air, which overcomes the pressure of the springs. On release the<br />
compressed air escapes <strong>and</strong> the pressure of the springs takes over, this forces first the<br />
upper <strong>and</strong> then the lower valve to close. The small space between the valves still exists<br />
<strong>and</strong> if avalve is not properly sealed liquid will flow through the space <strong>and</strong> out of the<br />
assembly through apipe. Leaking is thus detected. These systems are now favoured with<br />
the desire to lower manpower costs in breweries, but are only of use if closely examined<br />
frequently!<br />
Carbon dioxide produced during fermentation must be removed from the vessel<br />
irrespective of whether this gas is collected for further use (Section 15.4.3). This gas<br />
pressure must, therefore, be released <strong>and</strong> the vessel must also be protected against the<br />
development of avacuum. CIP fluids must be safely <strong>and</strong> effectively introduced into the<br />
tank. For ease of manufacture <strong>and</strong> ease of operation it is now common <strong>practice</strong> to<br />
incorporate these fittings into atop plate of about 1m(3ft.) diameter. If the vessels are<br />
st<strong>and</strong>ing in the open air this top plate must be protected against the weather.<br />
Pressure relief is usually by asimple weight-operated valve (Fig. 14.3), although<br />
spring controlled systems are sometimes used. Excess pressure can develop from vessel<br />
filling, <strong>and</strong> expansion of fermentation gases when hot CIP is used. A reduction in<br />
pressure can occur on emptying the vessel, from the reaction of carbon dioxide with<br />
causticdetergents <strong>and</strong>whencold liquid entersthetank after hot cleaning.Hot cleaningof<br />
fermenting vessels is always dangerous to the integrity of the tank <strong>and</strong>, if possible, it is<br />
recommended that fermenters are cleaned at temperatures of
adual planar system where the nozzles rotate on the head whilst the whole head rotates<br />
on the supply pipe (see Fig. 14.6 on page 525).<br />
Insulation<br />
Fermenters require insulation. Outside tanks are insulated against ambient conditions,<br />
which obviously vary considerably throughout the world. Indoor tanks will also require<br />
some insulation to lessen the dem<strong>and</strong>s on the temperature control system. Insulation is<br />
best added to the tank during manufacture for vessels up to asize of around 2000hl<br />
(1200imp. brl).<br />
Insulation materials usually contain chloride ions <strong>and</strong> so achloride inhibiting layer<br />
must be applied to the tank to protect the stainless steel prior to the application of the<br />
insulation. There is a wide range of insulation materials available <strong>and</strong> individual<br />
companiesoftenhaveverydistinctpreferences.Thebasicchoiceisbetweenpolyurethane<br />
foam <strong>and</strong> phenolic foam. Phenolic foam is less flammable although polyurethane can<br />
incorporate fire retardants, but these contain chloride. Inorganic materials such as glass<br />
fibre or mineral wool are completely non-flammable but are less good insulators. The<br />
thickness of the insulation used also varies widely. The capital cost of additional<br />
insulation is small in relation to the total cost of the vessel <strong>and</strong> so it is foolish to<br />
compromise here. Polyurethane foams are often applied in thicknesses of 100 to 150mm<br />
(4 to 6in.), whereas phenolic foams have been claimed to be effective at 75mm (3 in.)<br />
(Anderson, et al., 2000). Some brewers will specify amaximum permitted temperature<br />
rise in, say, summer conditions of
fermentations. For higher-gravity fermentations (say >12ëP, 1048ëSacch) it is now<br />
common to use oxygen instead of air. Concentrations of up to 30mg/l are possible but 12<br />
to 20mg/l are normal. Over oxygenation is sometimes thought to be impossible because<br />
oxygen is so rapidly utilized by the yeast. However, with some strains of yeast excess<br />
oxygen in the wort can lead to overgrowth <strong>and</strong> lower ethanol yields. It is therefore good<br />
<strong>practice</strong> to match the oxygen requirement of each particular yeast strain (Chapter 13).<br />
Oxygen is normally added after wort cooling but before the yeast. Anumber of types<br />
of injection system are in use. Sintered ceramic c<strong>and</strong>les can cause the injection of avery<br />
fine stream of bubbles to achieve intimate mixing but these systems are difficult to clean.<br />
Aventuripipeisveryeffective.Airoroxygenisintroducedjustbeforetheconstrictionin<br />
the pipe <strong>and</strong> then mixes thoroughly with the wort in the turbulent flow resulting from the<br />
subsequent increase in diameter of the pipe (see also Chapter 10).<br />
Yeast for pitching has normally been derived from previous fermentations <strong>and</strong><br />
maintained in yeast storage vessels as a slurry at < 5 ëC (41 ëF). The essence of successful<br />
pitching is the measurement of the quantity of the yeast to be pitched into the wort<br />
(Chapter 13). This can be achieved by volume, mass, or weight but these methods rely on<br />
the slurry being of constant composition <strong>and</strong> do not compensate for yeast viability.<br />
Methods are now available for making these compensations.<br />
· Volumetric methods. These methods are simple, cheap <strong>and</strong> can be successfully<br />
carried out by relatively unskilled personnel. A calibrated vessel can be used to<br />
measure the volume or more effectively a volumetric flow meter such as a magnetic<br />
flow meter. Peristaltic pumps have also been used. Account must be taken of the<br />
viability <strong>and</strong> concentration of the yeast to achieve good results. The method is affected<br />
by carbon dioxide gas trapped in the yeast slurry. This cannot be assessed <strong>and</strong> hence<br />
markedly different carbon dioxide concentrations in different slurries will affect the<br />
amount of yeast pitched into the wort.<br />
· Mass methods. The problem of trapped carbon dioxide can be avoided by the use of a<br />
mass meter, e.g., of the Coriolis type. Again account must be taken of the viability <strong>and</strong><br />
concentration of yeast in the slurry. Methods of this type are gaining use in breweries<br />
often in conjunction with instrumental systems to measure yeast concentration.<br />
· Weight methods. This is a common <strong>and</strong> effective system. If used with yeast slurries<br />
the weight method requires the yeast storage vessels to be installed on load cells so the<br />
vessel contents can be weighed. The method is also commonly used with pressed yeast<br />
cake, which is subsequently slurried in chilled water prior to addition. The main<br />
problem with these methods is the reliability of the load cells that do not work well in<br />
the conditions of the yeast storage area (wet <strong>and</strong> cold!). As with other methods account<br />
must be taken of viability <strong>and</strong> concentration of the yeast.<br />
As simply applied these methods suffer from the inherent variability of yeast slurries<br />
in terms of viability <strong>and</strong> concentration. The number of cells in a yeast slurry can be<br />
estimated in a laboratory using a Coulter counter. This figure can then be used to estimate<br />
the volume of slurry needed to give a particular yeast count in the wort. The Coulter<br />
counter will not distinguish between live <strong>and</strong> dead cells. To do this many brewers will<br />
apply a compensation factor to the amount of slurry to be pitched. Viability will be<br />
assessed by the methylene blue staining method. In this method dead cells stain blue <strong>and</strong><br />
viable cells remain colourless (Chapter 13). The method has, however, been considered<br />
unreliable for viabilities of < 85% (Institute of <strong>Brewing</strong>, Analysis Committee, 1971).<br />
There have been recent attempts to improve methods of determining concentration <strong>and</strong><br />
viability of slurries by using new technologies. There has also been the attempt to link<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
these methods into automatic systems of yeast pitching in breweries to avoid laboratory<br />
involvement <strong>and</strong> devolve the control of the process to brewery personnel. The objective<br />
here has been to improve the consistency of brewery fermentations <strong>and</strong> achieve more<br />
predictable attenuation <strong>and</strong> flavour volatile production. Askid-mounted instrument has<br />
been described (Teass, 2000) that will provide instantaneous cell counts in the 0to 2<br />
billion cells/ml range. Results obtained with the instrument correlated well with<br />
laboratory measurements. The volume to be pitched is controlled by aflow meter but<br />
corrections for viability still have to be made. Other instruments purport to measure<br />
viablecellconcentration.Abiomasssensormeasures thedielectricalpermittivityofyeast<br />
cell suspensions. This system utilizes aradio-frequency signal to create an electrical field<br />
through which the yeast cell suspension can flow. The yeast slurry acts as adielectric in<br />
that the electric field gives rise to no net flow of electric charge but only to a<br />
displacement of that charge. Areading of the displacement angle can be correlated with<br />
the concentration of intact cells. The assumption is made that intact cells are viable cells.<br />
On-line <strong>and</strong> off-line instruments have been developed using this principle (Carvell et al.,<br />
1998). The sensor can be calibrated to different yeast strains used for different brewery<br />
fermentations.<br />
The actual pitching rate used varies considerably between breweries <strong>and</strong> rates of 5to<br />
20millioncells/ml ofwort are common depending onthe specificgravity ofthe wort. An<br />
optimum level is considered to be 10 to 12 million cells/ml <strong>and</strong> this should result in a<br />
reproduction rate for lager yeast of 3to 5times (Stewart <strong>and</strong> Russell, 1998).<br />
Temperature control<br />
The heat output during fermentation has been discussed (Section 14.2.3). It follows that<br />
for reproducible lager fermentation in abrewery this excess heat must be removed by the<br />
cooling system so that the temperature of the fermenting wort can be controlled to a<br />
chosen profile. The beer must also be cooled at the end of fermentation to achieve the<br />
maturation temperature (Chapter 15) <strong>and</strong> to help the sedimentation of the yeast.<br />
All three methods of known heat transfer can affect brewery fermentations.<br />
Conduction occurs when heat is transferred through the vessel wall, e.g., from the<br />
insulation. Radiation can occur from, say, direct sunlight on the vessel surface. But the<br />
most important factor to deal with is convection arising from movements within the<br />
fermenting liquid itself. Convection can be caused by density gradients arising from<br />
unequal temperature distribution in the vessel. These convection currents are enhanced by<br />
the movement of the liquid caused by the natural evolution of carbon dioxide bubbles<br />
during the fermentation of the wort sugars. A complex system thus operates in the<br />
fermenter which can, however, be subjected to basic physical analysis. The rate of heat<br />
transfer can be expressed in an equation derived from Fourier's Law:<br />
Q ˆ U.A. t<br />
Where Q is the rate of heat flow or conductivity, W, measured in calories/second, A is the<br />
area of heat transfer in m 2 , U is the heat transfer coefficient in W/m 2 /ëC <strong>and</strong> t is the<br />
temperature difference in ëC.<br />
For a fermenting vessel the term A is the area of the wall of the vessel subject to the<br />
temperature difference (Andrews, 1997). This could be the whole of the vessel area if the<br />
effects of insulation are being considered or the area exposed to the cooling jackets if<br />
considering the effect of coolant. U is difficult to calculate <strong>and</strong> is dependent on a<br />
combination of factors relating to fluid viscosity <strong>and</strong> density <strong>and</strong> conductivity <strong>and</strong> fluid<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Density (g/cm 3 )<br />
1.000<br />
0.9999<br />
0.9998<br />
–2 0 2 4 6 8<br />
Ice starts<br />
forming<br />
Point of maximum<br />
density of beer<br />
Direction of cooling<br />
convection currents<br />
Temperature (°C)<br />
Point of maximum density<br />
of water, 3.98°C<br />
Fig. 14.4 Change in density of fermenting beer with change in temperature (Barnes, 2001).<br />
velocity at the vessel wall. Extraneous materials on the vessel wall, e.g., fouling by yeast<br />
deposits will also affect this fluid velocity <strong>and</strong> conductivity.<br />
If a temperature difference is now established between the fermenting beer <strong>and</strong> the<br />
vessel wall by means of a coolant, the beer adjacent to the vessel wall (sometimes called<br />
a beer film) will move downwards as the wall temperature will be lower than the beer<br />
temperature. A temperature profile will thus be set up between layers in the beer <strong>and</strong> the<br />
coolant. This will result in the establishment of density gradients further causing<br />
convection currents.<br />
An added factor to be considered in the temperature distribution in fermenters is the<br />
inversion temperature (Fig. 14.4). Water is most dense at 4 ëC (39.2 ëF). Water warmer<br />
than 4 ëC will rise <strong>and</strong> so on cooling a large tank, cold water (or fermenting beer) will<br />
initially flow down the tank wall until it reaches 4 ëC (the inversion temperature) when<br />
the flow pattern will reverse <strong>and</strong> the fluid will rise. The temperature of maximum density<br />
of beer is affected by alcohol content <strong>and</strong> extract. The temperature of maximum density<br />
of a 12 ëP (1048) beer is about 2.5 ëC (36.5 ëF) whereas a 16 ëP (1064) beer would have a<br />
maximum density at 1 ëC (33.8 ëF); lower gravity beers will obviously be closer to 4 ëC at<br />
maximum density.<br />
It is a real practical challenge to achieve a uniform temperature distribution in a<br />
cylindroconical fermenting vessel by using cooling jackets on the vessel <strong>and</strong> cone walls.<br />
We are dealing with the efficiencies of heat transfer from the beer film adjacent to the<br />
vessel wall into the bulk of the beer <strong>and</strong> heat transfer to the coolant from the beer film.<br />
There will also be the effect of the thickness of the vessel wall itself although as this is<br />
usually about 5 mm (approx 0.2 in.) the thermal conductivity is very high (4,000 W/m 2 /<br />
ëC). The driving force in temperature distribution is the convection currents. Having a<br />
large temperature difference between the coolant <strong>and</strong> the vessel wall will increase the rate<br />
of heat transfer <strong>and</strong> potentially improve temperature uniformity in the vessel. However,<br />
this is limited as freezing of the beer on the vessel wall must be avoided. Sophisticated<br />
systems have employed variable temperature of coolant using a lower temperature during<br />
fermentation than at the end <strong>and</strong> so avoiding freezing, however, these systems are too<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
complex to operate in most breweries <strong>and</strong> the temperature of coolant chosen is usually a<br />
compromise choice at about 5 ëC (23 ëF) to provide optimum cooling without freezing.<br />
Good thermal conductivity in the beer film adjacent to the vessel wall is to be<br />
encouraged. The strength of convection currents will theoretically be increased by a large<br />
temperature difference between the bulk of the beer <strong>and</strong> the film <strong>and</strong> the release of carbon<br />
dioxide creating turbulence at the peak of fermentation. But at the end of fermentation,<br />
movement in the vessel is again solely as a result of density differences which, at the<br />
temperature of maximum density will be at a minimum. A typical U value during<br />
fermentation for the beer side film transfer to the beer bulk would be 400 W/m 2 / ëC but<br />
this will be less when carbon dioxide production declines.<br />
Heat transfer to the coolant is easier to assess than heat transfer from the beer film to<br />
the bulk of the beer. A high velocity of coolant (1 m/s) to create turbulence is the most<br />
important factor in efficient heat transfer. A low viscosity is also favourable which is<br />
often the reason to prefer ethanol in the form of industrial methylated spirit (IMS) to<br />
propylene glycol:<br />
Viscosity of IMS, 7.5 cP at 3 to 5 C (27 to 23 F†<br />
Viscosity of propylene glycol, 18 cP at 3 to 5 C (27 to 23 F)<br />
A heat transfer coefficient of 2340 W/m 2 /ëC is achievable with IMS <strong>and</strong> about 1900 with<br />
propylene glycol. Ammonia is very effective as a primary refrigerant used as a<br />
compressed gas <strong>and</strong> of course has zero viscosity as far as these comparisons are<br />
concerned <strong>and</strong> can create high turbulence during evaporation resulting in a heat transfer<br />
coefficient of 4,000. The overall heat transfer coefficient therefore must be calculated<br />
from the sum of the individual coefficients:<br />
1=U ˆ 1=Ubeer side ‡ 1=Ucoolant side ‡ 1=Uwall resistance ‡ 1=Ufouling<br />
This equation includes a contribution of the effect of fouling on heat transfer on both the<br />
beer side <strong>and</strong> coolant side of the system. In <strong>practice</strong>, with secondary refrigerants such as<br />
IMS there is very little fouling on the coolant side. This is more of a factor in direct<br />
expansion systems with a gas like ammonia, which needs a lubricating oil on internal<br />
surfaces of the cooling jackets.<br />
There have been few reports on investigations of temperature distribution in large<br />
tanks in the last 25 years (Andrews, 1997) <strong>and</strong> many of the design proposals of brewery<br />
engineers relate to the theory discussed above <strong>and</strong> experimental investigations of the<br />
1970s (Maule, 1976, 1986). From these investigations <strong>and</strong> unpublished results (Barnes,<br />
2001) it is clear that the natural convection currents are inadequate to provide totally<br />
uniform cooling irrespective of the position of the cooling jackets on the vessel or the<br />
nature of the coolant used. Beer in the upper volume of a tank may hardly change its<br />
temperature throughout a cooling regime. Temperature probes in the lower zones of the<br />
tank may indicate that temperature control is being achieved but this is often not the case<br />
for all the beer in the tank. This is clearly unsatisfactory <strong>and</strong> can lead to inconsistencies in<br />
flavour volatile production <strong>and</strong> attenuation.<br />
This has prompted experiments in agitation of the contents of the vessel. Small amounts<br />
of a gas such as air or carbon dioxide can be introduced to the base of the vessel for this<br />
purpose. When carbon dioxide was injected through a sintered stainless steel c<strong>and</strong>le at 100g/<br />
min. for four minutes (Lemer et al., 1991) improvements in the uniformity of temperature<br />
<strong>and</strong> subsequent cooling were achieved. Mechanical rousers are in use in several breweries<br />
but they are difficult to maintain <strong>and</strong> present a potential source of microbial infection. A<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
novel proposal (Andrews, 1997) is to construct split cooling jackets so that only half of the<br />
circumference is providing cooling as the temperature is approaching that of maximum<br />
density, i.e., that at which flow inversion occurs. This will cause an increase in heat transfer<br />
coefficient, as fluid movement would be maintained. Temperature gradients would then be<br />
generatedhorizontally<strong>and</strong>notverticallythusavoidinglayeringinthetank.Agitationduring<br />
primary fermentation is disliked by some brewers because of the consequent increase in<br />
fermentation rate <strong>and</strong> increased organic acid production <strong>and</strong> lower beer pH value. It is,<br />
however, important to pay attention to the uniformity of temperature distribution in large<br />
cylindroconical vessels for regularity of beer flavour development.<br />
In asecondary coolant system the coolant is circulated by pumps through the cooling<br />
jacketsonthe vessels.The coolant itself iswarmed<strong>and</strong>isreturnedtoarefrigerationplant<br />
for further cooling. This system is thus `indirect'. In the refrigeration plant the secondary<br />
coolant is cooled by aprimary coolant, which evaporates <strong>and</strong> so takes heat from its<br />
environment. Chlorofluorocarbons (CFCs) have frequently been used for this purpose in<br />
breweries. The compounds are given an `R' number, e.g., R22, where the number relates<br />
tothenumberofcarbon,hydrogen<strong>and</strong>chlorineatoms,butthesesystemsofnomenclature<br />
are sometimes confused. These compounds have ozone-depleting effects <strong>and</strong> are<br />
gradually being replaced by ammonia according to internationally agreed protocols.<br />
However, ammonia is avery corrosive <strong>and</strong> toxic gas causing acute irritation of the<br />
respiratory pathway. It is also explosive when mixed with air at high temperatures. It<br />
follows that rigorous safety procedures must be in place in breweries where this gas is<br />
used in the refrigeration system.<br />
Ammonia gas can be used directly as the primary refrigerant in adirect expansion<br />
cooling system (Fig. 14.5). Liquid ammonia is pumped through the cooling jackets at a<br />
defined pressure <strong>and</strong> about 10% of the mass evaporates causing cooling. The mixture of<br />
gas <strong>and</strong> liquid is returned to a receiver. The gas is then re-compressed <strong>and</strong> condensed to a<br />
liquid <strong>and</strong> can be returned through the cooling jackets. Direct expansion systems<br />
consume between 35 <strong>and</strong> 45% less energy than indirect systems but the dangers of using<br />
ammonia remain <strong>and</strong> in many breweries indirect systems are in use with the trend to<br />
replace the primary refrigerant CFCs in these systems with ammonia.<br />
It is evident that control of temperature in the whole of a cylindroconical vessel is<br />
difficult. The position of the temperature probes is critical so that correct control of<br />
Ammonia vapour<br />
to compressors<br />
Pressure control signal<br />
to compressors<br />
Liquid ammonia<br />
from condenser<br />
Oil return to<br />
compressors<br />
LC<br />
PC<br />
Ammonia<br />
receiver<br />
Liquid<br />
ammonia<br />
For clarity, vessel is shown with a single jacket<br />
TC<br />
Ammonia liquid <strong>and</strong> vapour<br />
Fig. 14.5 Direct expansion cooling of a fermenting vessel; PC, pressure control; LC, level control;<br />
TC, temperature control (Anderson et al., 2000).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
coolant flow is achieved. The probe should be installed at aposition between 30 <strong>and</strong> 50%<br />
of the height from the base of the vessel <strong>and</strong> about 500mm (20in.) into the vessel from<br />
the wall to avoid influence from the temperature of the wall itself. In some systems a<br />
series of probes are installed <strong>and</strong> the overall temperature computed from these inputs.<br />
However, attention to ensuring the uniformity of temperature by underst<strong>and</strong>ing the<br />
gradients occurring in the vessel as aresult of convection cannot be too strongly stressed.<br />
Cleaning of vessels<br />
Cylindroconical vessels are now almost always cleaned by in-place cleaning systems<br />
(CIP).TheprinciplesofCIParediscussedinChapter17.CIPisexpensive<strong>and</strong>potentially<br />
a major contributor to brewery variable cost. The vessels have a high organic soil level<br />
(yeast deposits) <strong>and</strong> therefore respond to cleaning with hot caustic soda (1±2%).<br />
However, to save costs the bulk of the soil should be removed with jets of water in an<br />
impact system. Various designs of impact jet are available but the most effective for<br />
fermenters are the so-called multi-planar heads in which the nozzles on the head rotate in<br />
one plane whilst the head itself rotates at right-angles on the support pipe (Fig. 14.6).<br />
Liquid should not collect in the base of the vessel or effective cleaning of this area is lost.<br />
This liquid is usually removed by scavenge pumps which must be correctly sized <strong>and</strong><br />
have a capacity greater than the supply system <strong>and</strong> be capable of pumping with an empty<br />
supply system (self-priming). CIP cycles are now very carefully controlled by computer<br />
to optimize the use of materials <strong>and</strong> many variations can be set. A basic sequence would<br />
be based on:<br />
· Pre-rinse with cold recovered water (say 15 ëC, 59 ëF) with the objective of removing<br />
as much of the soil as possible. The soiled water would pass to the effluent system.<br />
Time 10±12 minutes.<br />
· Rinsing with hot (say 60 ëC, 110 ëF) caustic detergent (1±2% NaOH). Detergent will be<br />
re-circulated to a tank <strong>and</strong> will be replaced only when depleted (carbonate formed).<br />
Time 20±25 minutes.<br />
Sprayhead rotates<br />
in horizontal plane<br />
Sprayhead suspended<br />
by CIP supply pipe<br />
Nozzles rotate in<br />
vertical plane<br />
Fig. 14.6 A multi-planar CIP cleaning head (Anderson et al., 2000).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
· Post-rinsewithcleanwatertoremovedetergentresidues.Asterilantcouldbeaddedat<br />
this stage (Chapter 17). Water from this stage is recovered as pre-rinse. Time 10±12<br />
minutes.<br />
The importance of the pre-rinse stage in minimizing detergent use cannot be over<br />
emphasized. Hot caustic soda solutions must not be applied to tanks containing carbon<br />
dioxide or rapid dissolution of the gas <strong>and</strong> implosion of the tank will follow. Vacuum<br />
relief valves (section 14.3.2) will provide some protection against this but must not be<br />
relied on as acontrol measure.<br />
Acids can be used to clean fermenters. Over time beer stone (calcium carbonate/<br />
oxalate) will collect on vessel surfaces <strong>and</strong> on bends in pipe-work <strong>and</strong> sometimes in<br />
pumps. In this situation cleaning with 0.5±1.0% nitric acid or sometimes phosphoric acid<br />
is effective. Caustic detergent with asequesterant such as EDTA (ethylene diamine tetra<br />
acetate), which strips the calcium from the stainless steel <strong>and</strong> keeps it in solution is<br />
effective but is more expensive than acid. Acid cleaning systems possibly lower the<br />
overall time of cleaning compared to systems using caustic soda (Barnes, 2001).<br />
In fermentation vessel control systems CIP is often interlocked with aliquid level<br />
detector in the base of the tank. It is ruinous if aCIP sequence is started when beer is still<br />
in the tank. Double seat valves are therefore increasingly used in automated systems to<br />
provide security of operation <strong>and</strong> protection against leakage.<br />
14.4 Top fermentation systems<br />
Lager beers produced by bottom fermenting yeasts are by far the most widespread beer<br />
types throughout the world; consequently the bulk of development work on fermentation<br />
technologyhasbeenoncylindroconicalvesselsforbottomfermentingyeasts.Howeverin<br />
the UK <strong>and</strong> Irel<strong>and</strong>, ale <strong>and</strong> stout are the traditional beers <strong>and</strong> these are normally<br />
produced with strains of the top fermenting yeast Saccharomyces cerevisiae as are some<br />
Belgian <strong>and</strong> German beers. Fermentation systems developed to allow for the property of<br />
separating the yeast from the top of the fermenter at the end of fermentation. Traditional<br />
ale is cask conditioned (Chapters 21 <strong>and</strong> 23). Asecondary fermentation takes place inthe<br />
cask to provide condition to the beer. This ale is almost entirely produced by top<br />
fermentation. Ale can also be sold in kegs as a chilled <strong>and</strong> filtered product (Chapter 21).<br />
Some of this ale is produced by bottom fermentation in processes that are now difficult to<br />
distinguish from those of lager fermentation. Traditional ale brewers would regard<br />
producing ale by bottom fermentation with suspicion, but its proponents would cite the<br />
ease of separation of the yeast as the overriding issue.<br />
Removal of the yeast at the end of fermentation is traditionally by manual skimming<br />
or suction of the froth or `head' directly from the bulk of the fermented wort in the vessel.<br />
Developments of traditional systems have resulted in more ingenious methods for yeast<br />
separation. Examples of these methods are the Yorkshire Square system <strong>and</strong> the Burton<br />
Union system.<br />
14.4.1 Traditional top fermentation<br />
Vessels <strong>and</strong> rooms<br />
Traditional top fermentation utilizes a single vessel that is open to the atmosphere to<br />
facilitate yeast removal. The vessels were traditionally shallow (2 to 4 m, approx 6.5 to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fermenting<br />
beer<br />
Walkway<br />
Re-circulated air<br />
Louvres<br />
Air-cooling<br />
heat exchanger<br />
13ft. deep) <strong>and</strong> could be round, square or rectangular in shape (Fig. 14.7). Vessels have<br />
been made of many materials including wood, stone, slate, aluminium, cast iron, mild<br />
steel, copper, reinforced concrete <strong>and</strong> stainless steel. Vessels made of wood, cast iron,<br />
mild steel or concrete were usually lined with afurther material to assist cleaning.<br />
Linings were made of vitrified enamel, pitch, <strong>and</strong> various plastics (with or without the<br />
incorporation of fibreglass) <strong>and</strong> epoxy resins. Nearly all these linings had adverse<br />
features, mostly the possibility of tainting the beer. Some were fragile <strong>and</strong> needed<br />
frequent repair. Almost all top fermenting vessels built since the 1960s have been made<br />
of stainless steel, usually of type 304 (Section 14.3.2).<br />
Top fermenting vessels have traditionally been small (80 to 1000 hl; 50 to 550 imp.<br />
brl). Vessels have normally been grouped together in fermenting rooms. To lower the risk<br />
of microbial infection the surfaces in the fermenting rooms must be smooth <strong>and</strong> also,<br />
most importantly, accessible to easy cleaning. Walls are normally tiled or finished with<br />
polypropylene sheeting <strong>and</strong> floors are tiled or covered with asphalt or terrazzo. There<br />
must be a sufficient fall on the floor to allow for drainage <strong>and</strong> the drains must be<br />
constructed with traps to avoid odours. Condensation on ceilings is often a problem as<br />
condensate can fall into the fermenting beer. To avoid this the whole fermenting room is<br />
often air-conditioned.<br />
Normally a fermenting room has a false floor between the vessels usually about 600 to<br />
900 mm (23 to 36 in.) below the tops of the vessels <strong>and</strong> about 2.5 m (8 ft.) above the true<br />
floor. The space between the false floor <strong>and</strong> the true floor is often called a `shell' room.<br />
This space is utilized for circulating attemperated air <strong>and</strong> for mains <strong>and</strong> pipes. Air from<br />
above, heated by fermentation <strong>and</strong> containing carbon dioxide, can be aspirated into the<br />
shell room space, mixed with fresh air <strong>and</strong> cooled through a heat exchanger <strong>and</strong> then reintroduced<br />
above the false floor. Ideally the room should be kept at a temperature<br />
between 15 <strong>and</strong> 18 ëC (59 to 64 ëF).<br />
Fan<br />
Arrows – forced<br />
air circulation<br />
Fresh air<br />
Fig. 14.7 An open square fermenting vessel (after Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Carbondioxideevolvedfrom open fermentersisamajorhealthhazard.Themaximum<br />
concentration for exposure during an eight-hour shift is set at 0.5% (sometimes called the<br />
threshold limit value, TLV) in many countries including the UK <strong>and</strong> Germany. At 1to<br />
2% carbon dioxide, blood composition changes <strong>and</strong> oxygen access to the brain becomes<br />
restricted. At concentrations above 2% respiration rate increases in an attempt to<br />
compensate for the shortage of oxygen <strong>and</strong> dizziness is likely to be felt. Unconsciousness<br />
<strong>and</strong> death will follow at concentrations above 8%. Many accidents arising from exposure<br />
to carbon dioxide have occurred in brewing <strong>and</strong> accidents continue to occur. It is,<br />
therefore, extremely important to take great care when working with open fermenters.<br />
Carbon dioxide, being heavier than air, will collect in the walkways between vessels as it<br />
spills over from the fermenter. Asystem of positive air displacement must be used to<br />
ensure that this air is removed <strong>and</strong> if re-circulated (see above) must be enriched with<br />
fresh air to lower the carbon dioxide concentration to
topfermentedalesthanforlagers.Moremodern vessels,facilitatingeasiercleaning,have<br />
side wall cooling.<br />
The yeast is removed from the vessel by skimming; this is usually by suction but was<br />
once done manually with paddles. This removal of yeast may take place once, twice or<br />
three times depending on the conditions <strong>and</strong> the type of yeast in use. The first yeast crop<br />
is frequently dirty <strong>and</strong> will contain trub, particularly if adropping system is not in use.<br />
This crop will also be the least attenuative of that collected <strong>and</strong> it is often discarded. The<br />
following crops will be more attenuative <strong>and</strong> cleaner <strong>and</strong> these crops are usually retained<br />
for re-pitching although this relies on the experience <strong>and</strong> skill of the brewer. Yeast<br />
remaining on the base of the vessel (grounds) is normally discarded as this does not<br />
display the characteristics required for successful top fermentation, i.e., ayeast that<br />
attenuates rapidly <strong>and</strong> rises from the beer easily into the head at the completion of<br />
primary fermentation.<br />
Yeast removed from the fermenterby suction is held in avesselcalled ayeast back. In<br />
traditional systems, the yeast slurry would be processed through afilter press (Chapter<br />
15), to separate yeast from the entrained beer, `barm' ale, which would then be added<br />
back to the beer at racking usually after pasteurization. The yeast would be discharged<br />
from the press <strong>and</strong> held in trays or trucks in arefrigerator before re-use. The yeast can<br />
also be held without pressing, as aslurry at
Fish<br />
tail<br />
Pump for<br />
rousing<br />
For yeast<br />
removal<br />
Manhole<br />
Organ pipe<br />
Fig. 14.8 Yorkshire square fermenting vessel (Hough et al., 1982).<br />
flocculent. Beers of characteristic flavour are produced, which often at low gravities<br />
(
Enclosed vessels have been built (Griffin, 1996; Ogie, 1997) to allow full CIP, more<br />
effective microbiological control <strong>and</strong> the potential to collect carbon dioxide gas. Of major<br />
importance has been the novel method of yeast collection reducing the manual operations<br />
of the vessels. Vacuum yeast collection tanks are connected to a pair of yeast skimming<br />
points set in the top deck of the vessel. Suction operates through these points for a set<br />
period until most of the yeast has been removed. Spray lances are then used to deliver<br />
bursts of high pressure water across the deck to direct the remaining yeast to the<br />
skimming outlets. Each burst lasts about 10 seconds <strong>and</strong> is set to ensure the yeast<br />
concentration is sufficient for re-pitching <strong>and</strong> that water consumption is kept as low as<br />
possible. The base of the fermenter is fitted with pop-up spray jets (pintle valves), which<br />
operate to remove the grounds once the beer has been run to maturation or racking. All<br />
operations are remotely controlled by computer <strong>and</strong> the vessels are essentially automatic.<br />
Thus the traditional features of Yorkshire square fermentation are maintained with the<br />
minimum of manual involvement.<br />
14.4.3 Burton Union fermentation<br />
This system of fermentation is associated with the Burton-on-Trent area in Engl<strong>and</strong> <strong>and</strong><br />
was devised for the production of pale ales with powdery, i.e., non-flocculent yeasts.<br />
Wort is collected <strong>and</strong> pitched in a collecting vessel <strong>and</strong> then transferred at the peak of<br />
fermentation (36 to 40 hours) to the set of Union vessels. These vessels are oak casks of<br />
capacity 153 imp. gallons (7 hl) arranged in two adjacent rows of 12 vessels beneath an<br />
inclined, cooled `top' trough. The individual casks contain cooling coils. At the top of<br />
each cask is a swan-neck pipe, which can discharge into the trough (Fig. 14.9). Carbon<br />
dioxide gas carries yeast up through the swan-neck <strong>and</strong> it falls into the top trough. Yeast<br />
tends to sediment in the trough <strong>and</strong> beer collects at the lower end <strong>and</strong> is returned by side<br />
tubes to the casks. As fermentation completes virtually all the yeast has passed to the top<br />
trough. The beer in the casks is discharged into a second trough running below the row of<br />
casks <strong>and</strong> thence to racking backs. Beer transferred to racking is very clear <strong>and</strong> the yeast<br />
count is adjusted by the position of the outlet tap on each cask.<br />
a1 a2 s<br />
b<br />
c<br />
f<br />
d<br />
e<br />
s<br />
h h<br />
Fig. 14.9 Burton Union fermentation system (a 1 ) attemperator water (beer); (a 2 ) attemperator<br />
water (yeast); (b) side rod; (c) waste water; (d) top trough; (e) bottom trough; (f) feeder; (s) swannecks;<br />
(h) bottom tap; (m) side tap; (n) sample tap (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
s<br />
m<br />
n
As with Yorkshire squares, beer of characteristic flavour is produced <strong>and</strong> some brewers<br />
maintain that true Burton pale ale can only be made in this way. In 1990 one company<br />
invested in a new version of the Burton Union system constructed in stainless steel<br />
comprising four sets of 30 150 gallon Unions. Burton Union systems require alot of<br />
space for relatively low throughput <strong>and</strong> have ahigh capital cost. Labour involvement is<br />
high because, essentially, all the cleaning is manual. However, good quality yeast for repitching<br />
is invariably produced along with beer that is easy to fine (Chapter 15) <strong>and</strong> the<br />
opportunity is presented to exploit the method of production for enhanced marketing of the<br />
br<strong>and</strong>.<br />
14.5 Continuous fermentation<br />
Once the biochemistry of fermentation began to be understood towards the end of the<br />
nineteenth century (Pasteur, 1860; Chapter 13), some brewing scientists began to think<br />
they had limited control over the fermentation process. Surprisingly this stimulated a line<br />
of investigation on the continuous culture of yeast at high concentrations in which an<br />
active metabolic state was maintained <strong>and</strong> the problems of batch separation of yeast from<br />
beer avoided (DelbruÈck, 1892). Continuous fermentation of beer was thus attempted<br />
before 1900 <strong>and</strong> by 1906 at least five separate systems had been described, some of<br />
which were patented (Van Rijn, 1906). This patent (Van Rijn, 1906) is of some interest<br />
because in it are described processes involving the cascading of partly fermenting beer<br />
from one open tank to another, an idea that was picked up much later on in the 1950s<br />
(Wellhoener, 1954). These early systems of continuous fermentation did not achieve<br />
commercial success for reasons that are not entirely clear. It seems probable that the<br />
inability to avoid infection by micro-organisms was one reason <strong>and</strong> perhaps another was<br />
the resistance to change by the established brewers of the day.<br />
As sales of beer increased in the 1950s <strong>and</strong> 1960s there was renewed interest in the<br />
potential of continuous fermentation particularly in New Zeal<strong>and</strong>, Canada <strong>and</strong> the UK. At<br />
this time the brewing scientist was often represented at main board level in brewing<br />
companies <strong>and</strong> as such had considerable influence on company strategy. There was thus<br />
enhanced knowledge in companies of the brewing process, which could be coupled with<br />
developments in electronic process control equipment. As a result there was real hope for<br />
the commercial success of continuous systems with the advantages comprising:<br />
· lower capital cost<br />
· lower working capital because of less beer in process, as a result of faster throughput<br />
· lower product cost as a result of lower beer losses, more ethanol <strong>and</strong> less yeast<br />
· lower fixed costs because of less manpower as a result of less cleaning <strong>and</strong> automatic<br />
fermenter control.<br />
Companies attempted to exploit the technology <strong>and</strong> gave considerable support to their<br />
scientific proponents. By the early 1970s about 4% of UK beer production was by some<br />
form of continuous fermentation <strong>and</strong> the outst<strong>and</strong>ing technology of Morton Coutts was<br />
established in New Zeal<strong>and</strong> (Coutts, 1957). But many of the systems fell quickly out of<br />
use <strong>and</strong> by the late 1970s <strong>and</strong> early 1980s the only production scale continuous<br />
fermenters in use were <strong>and</strong> are in New Zeal<strong>and</strong>.<br />
The decline of continuous fermentation reflected the rise in the use of the large batch<br />
cylindroconical fermenting vessel as brewers responded to the `threat' of continuous<br />
fermentation by finding improved batch methods. Developments in the design <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
operation of these vessels in the 1970s were rapid. At the same time the influence of<br />
marketing was becoming manifest in brewing. After apeak in brewing sales in Europe in<br />
1979, economic recession in the early 1980s led to asearch for more innovative ways of<br />
exp<strong>and</strong>ing beer sales. This resulted in new product development often using different yeast<br />
strains, <strong>and</strong> the desire to present the public with agreater choice of drinks. Inevitably this<br />
meant that some of these drinks were produced in low volumes <strong>and</strong> some failed to achieve<br />
market success. Many continuous fermentation systems did not have the flexibility to<br />
h<strong>and</strong>le this type of production, which was much easier in batch processes. Acontinuous<br />
system was geared to produce ahigh volume of asingle type of product. It was also<br />
realized that to operate continuous systems successfully required highly trained personnel<br />
often working shifts <strong>and</strong> this led, paradoxically, to increased costs. The situation was<br />
perhaps different in the less developed market conditions of New Zeal<strong>and</strong> where Morton<br />
Coutts was able to foster <strong>and</strong> develop his system to commercial success over many years.<br />
Into the 21st century the use of immobilized yeast systems has dem<strong>and</strong>ed a<br />
reassessment of the value of continuous fermentation <strong>and</strong> this has been allied with huge<br />
developments in process control using computers, which were not available in the 1970s<br />
<strong>and</strong> 1980s. The physiological <strong>and</strong> biochemical behaviour of yeast in continuous<br />
fermentation is discussed in Chapter 13. In this Chapter practical aspects are considered.<br />
14.5.1 Early systems of continuous fermentation<br />
There are anumber of different ways of classifying continuous yeast culture <strong>and</strong> hence<br />
fermentation systems (Herbert, 1961; Chapter 13). Yeast can emerge in an `open' system<br />
with the beer or in a`closed' system when it is retained with the beer. In ahomogeneous<br />
system the yeast <strong>and</strong> fermenting beer is intimately mixed in astirred fermenter, whereas<br />
in heterogeneous systems there will be separation <strong>and</strong> concentration of the yeast away<br />
from the beer in avessel. The cascade open system of the series of inter-connecting<br />
vessels was described in 1954 (Wellhoener, 1954). Substrate (wort) entered the first<br />
vessel <strong>and</strong> product (beer) emerged from the last. Residence time was about 25 days <strong>and</strong><br />
the system was never fully adopted on the commercial scale. But some open systems did<br />
achieve limited production success around 1970.<br />
Stirred tank fermenters<br />
Asystem (Bishop, 1970) operated in four UK breweries achieving an output of 32,000hl<br />
per week ( 20,000imp. brl). There were two stirred fermenters in series <strong>and</strong> a<br />
sedimentation vessel for collecting the yeast (Fig. 14.10). Yeast was not recycled <strong>and</strong> the<br />
residence time was around 15 hours. Other UK breweries operated similar systems. The<br />
problems for further enhancement were as discussed above <strong>and</strong> to achieve outputs<br />
comparable to those of the cylindroconical vessels of the day (with afive day turn round<br />
time); the stirred fermenters would have to be increased sixfold in size. This was not<br />
commercially feasible.<br />
Tower fermenters<br />
Tower fermenters were developed by the APV Company in the 1960s <strong>and</strong> were used in<br />
breweriesinthe1970s(Portno,1973).Thisheterogeneoussinglevesselsystem(Fig.14.11)<br />
utilized the ability of flocculent yeast strains to sediment <strong>and</strong> so maintain a high<br />
concentration of cells in the system. Wort was pumped into the base of the vertical tube. The<br />
sedimentary yeast formed a plug in the base of the vessel <strong>and</strong> the wort permeated through the<br />
plug. Some of the yeast was carried up the tower by the flow of wort <strong>and</strong> the fermentation<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Wort in<br />
Pump<br />
Oxygen<br />
column<br />
Sterilizer<br />
Stirrer<br />
drive<br />
Stirrer<br />
drive<br />
Fob<br />
breaker<br />
Cooling<br />
coil<br />
Beer<br />
outlet<br />
Yeast out<br />
CO 2 outlet<br />
Sedimentation<br />
vessel<br />
Fig. 14.10 Stirred tank continuous fermentation, two stirred fermenters in series <strong>and</strong> a yeast<br />
sedimentation vessel (Bishop, 1970).<br />
b<br />
e<br />
a<br />
h<br />
c<br />
d<br />
b<br />
Head<br />
8.2 m<br />
Top<br />
7.0 m<br />
26.3 brl<br />
Middle<br />
3.9 m<br />
14 brl<br />
Bottom 1.2 m 3.25 brl<br />
Fig. 14.11 Continuous tower fermenter; (a) wort collection; (b) impeller pump; (c) flow meter; (d)<br />
control valve; (e) flash pasteurizer; (f) tower; (g) yeast separator; (h) beer receiver; (j) CO2<br />
collecting vessel (after Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
f<br />
g<br />
j
proceeded as the wort rose. The rate of wort injection was adjusted so that at the top of the<br />
tower the wort was fully fermented. The tower contained baffles to control the upward<br />
movement of the yeast, which can be too rapid as aresult of carbon dioxide evolution in<br />
large bubbles. Yeast fell from suspension at the top of the tower <strong>and</strong> an inclined chute near<br />
thebeeroutflowassistedthis.Thesystemwasnotadaptablefortheuseofpowderyyeast<strong>and</strong><br />
therefore could not be used for producing beers associated with these yeasts.<br />
It was desired to produce the same crop of yeast as with conventional fermentation so<br />
that these could be mixed if required. This required adjustments in wort composition as<br />
yeast in high concentration grows slowly <strong>and</strong> does not assimilate the normal amount of<br />
nitrogenous material from the wort. Worts with high levels of fermentable carbohydrate<br />
<strong>and</strong>lowlevelsofassimilablenitrogenwere usedtoyieldbeersofnormalcomposition.This<br />
in turn leads to complications <strong>and</strong> increased costs in the management of the system. There<br />
were further difficulties with abnormal production of acids, esters, higher alcohols <strong>and</strong><br />
diacetyl. Tower fermenters were also prone to infection with lactic acid bacteria requiring<br />
wort to be pasteurized before fermentation <strong>and</strong> so adding to cost. Residence times in the<br />
fermenter were as low as four hours for ale fermentations but like the stirred tank<br />
fermenters <strong>and</strong> for similar reasons, prolonged production performance was not achieved.<br />
14.5.2 The New Zeal<strong>and</strong> system<br />
In view of the difficulties described above it is aconsiderable achievement that the<br />
system developed by Morton Coutts (Coutts, 1957, 1958) remains acommercial success.<br />
This is attributable to sound basic design, <strong>and</strong> unfailing commitment <strong>and</strong> skill.<br />
Significantly,theNewZeal<strong>and</strong>method isanintegralpartofacontinuousbrewingsystem<br />
(Fig. 14.12) <strong>and</strong> it comprises:<br />
· ahold-up vessel<br />
· two stirred fermenters<br />
· ayeast separator.<br />
In the hold-up vessel the wort is aerated <strong>and</strong> mixed with yeast <strong>and</strong> beer recycled from the<br />
first fermenter. The residence time is about four hours <strong>and</strong> considerable yeast growth<br />
occurs. The volume of the hold-up tank is 230hl (140imp. brl) <strong>and</strong> the following<br />
fermenters are 1,637hl (1,000imp. brl) <strong>and</strong> 409hl (250imp. brl) respectively.<br />
Temperature is maintained at 15ëC (59ëF) <strong>and</strong> the throughput is 70hl per hour. Wort<br />
of 18ëP (72 ëSacch) is fed to the hold-up vessel <strong>and</strong> is diluted in the first fermenter to<br />
13ëP (52ëSacch). The specific gravity of the beer emerging is about 3ëP (12ëSacch).<br />
Total residence time is about 30 hours <strong>and</strong> continuous runs of up to one year have been<br />
achieved. The process hasundergone anumberof changes since its introduction (Stratton<br />
et al., 1994). These have focused on improved microbiological control <strong>and</strong> process<br />
automation, which have resulted in major efficiency gains. Several plants are now<br />
operating(in2003). This systemhas been asuccessbecausesoundbeerstruetothe br<strong>and</strong><br />
type have been consistently brewed <strong>and</strong> the system has not been required to produce<br />
different beers on an intermittent basis.<br />
14.5.3 Continuous primary fermentation with immobilized yeast<br />
Technology<br />
Afundamental principle of continuous culture of yeast is the maintenance of the yeast in<br />
high concentration in its substrate (Chapter 13). All continuous brewing fermentation<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Air<br />
From<br />
lauter<br />
tun<br />
Filtered wort storage<br />
1°C (34°F)<br />
Hold-up<br />
vessel<br />
To finished<br />
beer<br />
collection<br />
Intermediate<br />
boiling<br />
Hops<br />
vessel Hop<br />
strainer<br />
Wort<br />
boiler<br />
Hot trub<br />
Yeast for propagation<br />
Stabilizing vessel<br />
0°C (32°F)<br />
CF 1 CF 2<br />
Fermented beer return for pH reduction Yeast in suspension<br />
control<br />
Chilling<br />
CO2<br />
Sucrose<br />
priming<br />
Diatomaceous<br />
earth filter<br />
Hop Caramelized<br />
extract sugar Finings<br />
Fermented wort<br />
O.G. dilution<br />
Surplus yeast <strong>and</strong><br />
precipitated material<br />
Yeast<br />
separator<br />
Cooling<br />
<strong>and</strong><br />
chilling<br />
Fig. 14.12 New Zeal<strong>and</strong> system of continuous fermentation, (CF1) continuous fermenter 1, (CF2)<br />
continuous fermenter 2 (Coutts, 1957, 1958).<br />
systems must have ways of achieving this objective. However, yeast in high<br />
concentration grows slowly <strong>and</strong> the metabolic effects on the wort are different from<br />
those achieved by batch fermentation at normal yeast concentrations. This factor,<br />
together with the proneness of continuous systems to become infected with beer spoilage<br />
bacteria limited the commercial development of continuous fermentation. This situation<br />
has changed with the availability of immobilized cell technology that offers the<br />
possibility of real heterogeneous fermentation.<br />
Immobilizationinvolvesthephysicalconfinementorlocalizationofwhole intactyeast<br />
cells in afixed position whilst retaining normal viability <strong>and</strong> metabolic activity (Chapter<br />
13). The yeast cells are no longer free <strong>and</strong> therefore, in a fixed position, can be<br />
maintained in high concentration. The most common immobilization technique (Stewart<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>and</strong> Russell, 1998) is to trap the yeast in amatrix. This is usually apolymer that forms a<br />
gel around the cells. Many polymers have been used: alginate, carrageenan, agar, pectin,<br />
polyacrylamide <strong>and</strong> silica gel, etc. After trapping the cells on the polymer, growth in a<br />
nutrient medium takes place to ensure that the matrix is fully saturated with yeast cells.<br />
Themixtureisthen`gelled'intosheets<strong>and</strong>subsequentlycutintoparticlesofdesiredsize.<br />
Spherical beads of 0.3±3.0 mm in diameter are normally used. The matrix is porous<br />
enough to allow the free diffusion of substrates <strong>and</strong> fermentation products. The gel must<br />
have mechanical strength so that it does not split in operation <strong>and</strong> allow yeast cells to<br />
leave the matrix. This is often as aresult of the evolution of carbon dioxide gas bubbles<br />
from the entrapped cells. This was amajor problem in early immobilized yeast systems.<br />
Somesystemshaveusedimmobilizationonpre-formedporousornon-poroussupports<br />
such as kieselguhr, perlite, wood chips, PVC chips, glass fibres, stainless steel <strong>and</strong> silica.<br />
These systems reduce mass transfer problems in that the cells are in closer contact with<br />
the substrates, but they are more prone to physical disruption particularly by carbon<br />
dioxide gas bubbles. Like any other continuous system, to be commercially viable an<br />
immobilized yeast system must offer repeatable advantages over batch systems. This<br />
means considerable work to optimize immobilization procedures, <strong>and</strong> mass transfer to<br />
ensure consistentlyhigh rates of fermentation. Recent reportssuggest these aims can now<br />
be achieved <strong>and</strong> that immobilized yeast bioreactors are capable of achieving rapid<br />
reproducible primary fermentation of wort over continuous periods of many months.<br />
Initially immobilized systems were primarily used for accelerated maturation of beer<br />
(Chapter 15) with rapid removal of diacetyl but now primary fermentation is achievable.<br />
Operation<br />
Early attempts at primary fermentation were not successful. This was partly as a result of<br />
prejudice arising from the perceived failure of free cell systems. But enthusiasts<br />
continued to experiment <strong>and</strong> received some encouragement from reports from the fuel<br />
alcohol industry (Margaritas <strong>and</strong> Merchant, 1984) which suggested reduced costs <strong>and</strong><br />
operating consistency were achievable with immobilized yeast bioreactors. Mass transfer<br />
limits were the problem in early immobilized technology (Masschelein et al., 1985) <strong>and</strong> it<br />
proved difficult to produce beers directly flavour matched to beers from batch<br />
fermentations. This was partly owing to the use of worts with too low levels of free<br />
amino nitrogen (FAN) resulting in unbalanced concentrations of higher alcohols <strong>and</strong><br />
esters. There was also low oxygen tension in early immobilized systems which was<br />
advantageous for ethanol production but not for the balance of beer flavour volatiles.<br />
In a Japanese two-stage process a free cell stirred fermenter is followed by an<br />
immobilized yeast bioreactor (Inoue, 1995). Yeast growth occurs in the stirred vessel,<br />
which allows appropriate utilization of FAN similar to that of conventional batch<br />
fermentation. Full attenuation was then achieved in the anaerobic bioreactor where the<br />
yeast cells were trapped on porous ceramic beads. Centrifugation of the beer between the<br />
two stages was often carried out. The system was combined with a maturation column<br />
(Chapter 15) to complete beer production in three to five days. From a 20-litre pilot plant,<br />
scale up to 100 hl has been achieved which has operated for two years. Whilst the system<br />
was resistant to bacterial infection it did not achieve the efficiencies hoped for <strong>and</strong> it did<br />
not offer significant improvement over batch fermentation. Capital costs were higher than<br />
anticipated <strong>and</strong> so were revenue costs as a result of higher beer losses in centrifugation<br />
<strong>and</strong> increased energy usage.<br />
A novel approach to solving the problem of efficient mass transfer has been developed<br />
in Canada (Mensour et al., 1997). This system utilizes a gas lift draft tube bioreactor (Fig.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Stainless steel<br />
head plate<br />
Inoculated beads<br />
Sparging gas<br />
Plant wort<br />
‘Green’ beer<br />
Thermal jacket<br />
Bioreactor<br />
Draft tube<br />
Fig. 14.13 Gas lift draft tube bioreactor (Mensour et al., 1997).<br />
14.13) with carrageenan gel beads. Good mixing is achieved with minimum shear on the<br />
matrix; hence mass transfer is considerably improved. Amixture of air <strong>and</strong> carbon<br />
dioxide was used as the sparging gas. The proportion of air affected the flavour profile of<br />
the resulting beer <strong>and</strong> by taste panel evaluation air concentrations of between 2±5%<br />
yielded acceptable products though not precisely matched to conventionally fermented<br />
beer. Again the system can be combined with accelerated maturation to produce finished<br />
beer in two days.<br />
An alternative two-stage process has been developed in Belgium (Masschelein <strong>and</strong><br />
Andries,1996).Theimmobilizedyeastbioreactorprecedesacylindroconicalstirredtank.<br />
The yeast is immobilized in rods of porous sintered silicon carbide of length 900mm <strong>and</strong><br />
diameter 26mm. Fermenting beer is circulated through the internal channels of the<br />
carbiderods.Greenbeerisdrawnfromthetopofthereactortothestirredtankinwhichit<br />
can also be circulated from the tip of the cone to apoint on the cylinder just above the<br />
start of the cone. The fermentation in the stirred tank is completed by yeast cells, which<br />
have escaped from the immobilized bioreactor. Beers of acceptable flavour have been<br />
produced at rates of 135hl/hl of fermenter volume over acontinuous period of six<br />
months.<br />
Further developments have been reported from Finl<strong>and</strong> (Andersen et al., 1999;<br />
Pajunen et al., 2000) where there has been considerable progress on achieving<br />
satisfactory maturation with immobilized yeast (Chapter 15). This work has concentrated<br />
on solving the problems associated with temperature control <strong>and</strong> the disruptive effect of<br />
carbon dioxide gas bubbles on the stability of the matrix. A single reactor is used<br />
operating at a high enough pressure to keep carbon dioxide gas in solution. The gas is<br />
removed in a separate de-gassing unit <strong>and</strong> the beer is circulated around the fermentation/<br />
de-gassing loop to complete attenuation. The heat generated in fermentation is removed<br />
with a heat exchanger. The system has operated in pilot scale at 50 litres per hour to yield<br />
beer in 20 hours.<br />
Future<br />
Work on continuous fermentation with immobilized yeast bioreactors is characterized by<br />
the great innovation <strong>and</strong> enthusiasm of its devotees. Nevertheless, problems remain over<br />
the widespread adoption of the technique. Rapid fermentation tends to result in the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
production of higher levels of diacetyl than are normal in brewery fermentations. This<br />
diacetylcanberemovedusingimmobilizedyeasttechnology(Chapter15).Somebrewers<br />
view this as an extra, avoidable process <strong>and</strong> are deterred from using continuous rapid<br />
primary fermentation. This does not detract from using immobilized yeast for rapid<br />
diacetyl removal in maturation. The use of agenetically modified yeast could overcome<br />
this problem of increased diacetyl production but this is unlikely to achieve public<br />
acceptance for some time, if ever. The external addition of -acetolactate decarboxylase<br />
would also help but some brewers wish to avoid the use of added enzymes.<br />
There is also the problem of the diversity of br<strong>and</strong>s produced by some brewers. If the<br />
production of different beers requires the use of different yeasts then the inflexibility of<br />
continuous fermentation is manifest. Successful continuous fermentation requires the<br />
steady production of asingle br<strong>and</strong> with great consistency (Section 14.5.2). To achieve the<br />
undoubted potential benefits of continuous fermentation would require different br<strong>and</strong>s to<br />
be produced from varying wort composition or from treatments post-fermentation.<br />
Optimistic conclusions for the future were drawn in an EBC symposium on<br />
immobilized yeast reported in 1997 (Linko et al., 1997), nevertheless in the same year<br />
99.9% of the World's beer was produced in batch fermentation systems (Masschelein,<br />
1997).<br />
14.6 Fermentation control systems<br />
The control of temperature in fermentation has received much attention (Section 14.3.3).<br />
These control systems are reactive to the exothermic nature of fermentation <strong>and</strong> are now<br />
highly developed. However, until recently little attention has been paid to the control of<br />
brewery fermentation in relation to the chemical changes which are taking place. These<br />
changesinclude thefallinspecificgravity<strong>and</strong>inpHvalueofthewort<strong>and</strong>theproduction<br />
of ethanol <strong>and</strong> carbon dioxide.<br />
In the general literature of biotechnology the control of industrial fermentations to<br />
yield many products is considered in depth. Brewers have not always made best use of<br />
this information, but any control system, no matter how innovative, must justify itself in<br />
terms of its cost effectiveness for the production of potable beer (Moll, 1983; Dauod,<br />
1987).<br />
14.6.1 Specific gravity changes<br />
The most widely investigated methods of fermentation control have been based on the<br />
fall in specific gravity as wort is fermented to yield beer. In its most basic form this<br />
method involves taking arepresentative sample of wort <strong>and</strong> measuring its attenuation<br />
limit using, e.g., EBC Analytica IV method. This can be done throughout fermentation to<br />
construct atime-course picture of attenuation during the fermentation process (Section<br />
14.2.2). The system can be automated such that the density of the wort is determined by<br />
direct methods (Fig. 14.14); this is, of course, asystem of automated measurement rather<br />
than control. Wort density could also be measured by refractometry or by the use of the<br />
oscillating U-tube. These methods have been developed into control systems when the<br />
fall in gravity has been used to cause an external change such as the onset of fermenter<br />
cooling. This has the potential of earlier use of cooling <strong>and</strong> hence savings in process time.<br />
In large cylindroconical fermenters pressure difference measurements have been used<br />
to monitor the density change. As sugars are fermented to ethanol <strong>and</strong> to carbon dioxide<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
CIP<br />
supply<br />
FV<br />
CIP<br />
supply<br />
Sample<br />
pump<br />
Specific gravity<br />
signal to control<br />
system<br />
Mass<br />
flowmeter<br />
so the density drops <strong>and</strong> the pressure in the fermenter will change in proportion. Pressure<br />
transducers are used to measure this change <strong>and</strong> by linking to acomputer the extent of<br />
fermentation can be followed. This technique has also been applied in new Yorkshire<br />
square fermenters (Griffin, 1996).<br />
14.6.2 Other methods<br />
It was demonstrated in the 1970s (Alford, 1976) that the partial pressure of carbon<br />
dioxide in exhaust fermenter gas was an excellent approximation to the partial pressure<br />
exertedbydissolvedcarbondioxideintheliquidinthefermenter.Thisprovidedthebasis<br />
for apotential control system to assess the extent of fermentation by measurement of<br />
carbon dioxide partial pressure in the exhaust gas. Acontrol method relating to the<br />
production of carbon dioxide has been described (Dauod et al., 1989), but has not<br />
achieved widespread adoption.<br />
The use of dielectrical permittivity for the control of yeast pitching rate has already<br />
been discussed (Section 14.3.3 <strong>and</strong> Chapter 13). The commercially available probe can<br />
also be used to detect viable yeast cells in streams of yeast emerging from the cone of<br />
MF<br />
ø25mm line size<br />
Fig. 14.14 Specific gravity monitoring system using a sample loop; the product sampled is<br />
fermenting wort in the specific gravity range 1000±1100 ëSacch, a signal is sent from the mass<br />
flowmeter to the control system, sample pump must operate in both directions to allow for CIP;<br />
(FV), fermenting vessel (Barnes, 2001).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
cylindroconical vessels during cropping (Boulton <strong>and</strong> Clutterbuck, 1993). This biomass<br />
probe can be used in-line to automatically control yeast removal from the fermenter at the<br />
end of fermentation. The proportion of the yeast crop most suitable for re-pitching can<br />
therefore be retained. This technique could also be of direct financial benefit in allowing<br />
the reduction of yeast storage capacity.<br />
Methods of fermentation control involving the measurement of pH value are also<br />
available (Barnes, 2001). A sample is withdrawn from the fermenter the pH is measured<br />
<strong>and</strong>, if it is too high, oxygen can be introduced to the fermenter automatically. This can<br />
be further linked into modulation of the coolant control valves to direct the temperature in<br />
the fermenting vessel. Possibly, future developments in fermentation control will focus<br />
on assimilating a number of measurements: density, carbon dioxide production, ethanol<br />
production, pH value into a computer <strong>and</strong> causing outputs to achieve automatically the<br />
desired time course of the fermentation. This will mainly result in improvements in the<br />
control of cooling by better modulation of coolant valves, with consequent savings in<br />
energy. The capital cost of such systems will, however, be high <strong>and</strong> anticipated paybacks<br />
must be carefully assessed before proceeding with installation.<br />
14.7 Summary<br />
The major developments in fermentation technology in the latter part of the 20th century<br />
<strong>and</strong> into the present day have focused on the cylindroconical fermenting vessel, which<br />
has been operated in batch sizes of up to 6,000 hl (3,600 imp. brl). The development of<br />
the vessel has reflected the increased dominance in the brewing world of beers produced<br />
by bottom fermenting yeasts. Recent developments of the cylindroconical vessel have<br />
included improvements in yeast pitching control to maximize the pitching of viable yeast<br />
<strong>and</strong> improved temperature control <strong>and</strong> cleaning. Overall fermentation control systems,<br />
where changes in the state of the fermenter can be effected automatically arising from the<br />
measurement of primary fermentation parameters such as density, are not in wide use.<br />
The difficulty of achieving cost effectiveness in this area provides encouragement for the<br />
proponents of continuous fermentation technologies.<br />
Continuous primary fermentation with free cell systems is difficult to exploit<br />
consistently <strong>and</strong> commercially. Large volumes of an established br<strong>and</strong> in a consistent<br />
market, <strong>and</strong> skill <strong>and</strong> enthusiasm of operators are needed for success. This has been<br />
achieved only in New Zeal<strong>and</strong>. Immobilized yeast technology would seem to be the most<br />
likely continuous system for widespread applicability. Extremely rapid fermentation is<br />
achievable (20 hours) <strong>and</strong> this can be allied to rapid maturation also with immobilized<br />
yeast. However, a problem remains, the perceived inflexibility of continuous systems <strong>and</strong><br />
inability to deal with a range of beer br<strong>and</strong>s requiring different yeasts. Nevertheless<br />
developments in the use of immobilized yeast for continuous primary fermentation will<br />
occur, as there is real potential for major reduction in brewery variable cost when this<br />
technique succeeds.<br />
14.8 References<br />
ALFORD, J. S. (1976) Can. J. Microbiol. 22 (1), 52.<br />
ANDERSEN, K., BERGIN, J., RANTA, B. <strong>and</strong> VILJAVA, T. (1999) Proc. 27th Congr. Eur. Brew. Conv., Cannes,<br />
771.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ANDERSON, R. G., BRITES SANCHES, A., DEVREUX, A., DUE, J., HAMMOND, J., MARTIN, P. A., OLIVER-DAMEN,<br />
B. <strong>and</strong> SMITH, I. B. (2000) Fermentation <strong>and</strong> Maturation Manual of Good Practice, European<br />
Brewery Convention, Zoeterwoude, The Netherl<strong>and</strong>s.<br />
ANDREWS, J. M. H. (1997) Ferment, 10, 309.<br />
BARNES, Z. C. (2001) Personal Communication.<br />
BISHOP, L. R. (1970) J. Inst. <strong>Brewing</strong>, 76, 173.<br />
BOULTON, C. A. <strong>and</strong> CLUTTERBUCK, V. J. (1993) Proc. 24th Congr. Eur. Brew. Conv, Oslo, 509.<br />
CARVELL, J., ODDI, L. <strong>and</strong> HARDING, C. (1998) Proc. 25th Conv. Inst. Brew. (Asia Pacific Section), Perth,<br />
201.<br />
CHRISTIAN, A. H. R. (1959) Brewers' Guard. 88, 43.<br />
COUTTS, M. W. (1957) British Patents, 872391±872400.<br />
COUTTS, M. W. (1958) Australian Patent, AU 216618.<br />
DAUOD, I. (1987) Brewers' Guard. 116 (6), 14.<br />
DAUOD, I., DYSON, R., IRVINE, J. <strong>and</strong> CUTHBERTSON, R. C. (1989) Proc. 22nd Congr. Eur. Brew. Conv.,<br />
Zurich, 323.<br />
DELBRUÈ CK, M. (1892) Wochensch. Brau., 9, 695.<br />
FRICKER, R. (1978) Brewers' Guard. 107, 28.<br />
GRIFFIN, S. R. (1996) Brewers' Guard. 125 (8), 12.<br />
HERBERT, D. (1961) Continuous Culture of Microorganisms, SCI Monograph, SCI London, 12, 21.<br />
HOGGAN, J. (1977) J. Inst. <strong>Brewing</strong>, 83, 133.<br />
HOUGH, J. S., BRIGGS, D. E., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1982) Malting <strong>and</strong> <strong>Brewing</strong> Science 2nd<br />
Edition Volume 2, Aspen, Gaithersburg, Maryl<strong>and</strong>.<br />
INOUE, T. (1995) Proc. 25th Congr. Eur. Brew. Conv., Brussels, 25.<br />
INSTITUTE OF BREWING ANALYSIS COMMITTEE (1971) J. Inst. <strong>Brewing</strong>. 77, 181.<br />
KNUDSEN, F. B. <strong>and</strong> VACANO, N. L. (1972) Brewers' Digest, 47 (7), 68.<br />
LEMER, C., TAEYMANS, D. <strong>and</strong> MASSCHELEIN, C. A. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon,<br />
329.<br />
LINKO, M., VIRKAJAVI, I., POHJALA, N., LINBORG, K., KRONLOF, J. <strong>and</strong> PAJUNEN, E. (1997) Proc. 26th Congr.<br />
Eur. Brew. Conv., Maastricht, 385.<br />
MAHLER, H. R. <strong>and</strong> CORDES, E. H. (1969) Biological Chemistry, New York, Harper-Row.<br />
MARGARITAS, A. <strong>and</strong> MERCHANT, F. (1984) Crit. Reviews in Biotehnology, 1, 339.<br />
MASSCHELEIN, C. A. (1997) J. Inst. <strong>Brewing</strong>, 103, 103.<br />
MASSCHELEIN, C. A. <strong>and</strong> ANDRIES, M. (1996) Brew. Distill. Internat. 27 (7), 16.<br />
MASSCHELEIN, C. A., CARTIER, A., RAMOS-JEUNNEHOMME, C. <strong>and</strong> ABE, L. (1985) Proc. 20th Congr. Eur.<br />
Brew. Conv., Helsinki, 339.<br />
MAULE, D. R. (1976) The Brewer, 62 (5).<br />
MAULE, D. R. (1977) The Brewer, 63 (6), 204.<br />
MAULE, D. R. (1986) J. Inst. <strong>Brewing</strong>, 92, 137.<br />
MENSOUR, N. A., MARGARITAS, A., BRIENS, C. L., PILKINGTON, H. <strong>and</strong> RUSSELL, I. (1997) J. Inst. <strong>Brewing</strong>,<br />
103, 363.<br />
MOLL, M. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, 272.<br />
NATHAN, L. (1930) J. Inst. <strong>Brewing</strong>, 36, 538.<br />
OGIE, P. J. (1997) Brew. Distill. Internat, 28 (6), 32.<br />
PAJUNEN, E., RANTA, B., ANDERSEN, K., LOMMI, H., VILJAVA, T., BERGIN, J. <strong>and</strong> GUERCIA, H. (2000) Proc.<br />
27th Conv. Inst. Brew. (Asia Pacific Section), Singapore, 91.<br />
PASTEUR, L. (1860) Annales de Chimie et de Physique, 3 e Serie. 58, 323.<br />
PASTEUR, L. (1876) Etudes sur la bieÁre, Paris, Gauthier-Villers.<br />
PORTNO, A. D. (1973) Brewers' Guard. 102 (7), 33.<br />
STEWART, G. G. <strong>and</strong> RUSSELL, I. (1998) Brewer's Yeast, Institute of <strong>Brewing</strong>, London.<br />
STRATTON, M. K., CAMPBELL, S. J. <strong>and</strong> BANKS, D. J. (1994) Proc. 23rd Conv. Inst. Brew. (Asia Pacific<br />
Section), Sydney, 196.<br />
TEASS, H. A. (2000) Tech. Quart. MBAA, 37 (1), 37.<br />
VAN RIJN, L. A. (1906) British Patent, 18045.<br />
WELLHOENER, H. J. (1954) Brauwelt, 94, 44, 624.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
15<br />
Beer maturation <strong>and</strong> treatments<br />
15.1 Introduction<br />
Beer, at the completion of primary fermentation is said to be `green'. It contains little<br />
entrained carbon dioxide, it is hazy <strong>and</strong> its taste <strong>and</strong> aroma are inferior to beer that is<br />
ready for sale. In order to refine green beer it must be matured or conditioned. This<br />
maturation processtakesplacein closed containers inthe brewery <strong>and</strong> beer treated inthis<br />
way is called brewery conditioned beer. This is most of the beer produced <strong>and</strong> sold in the<br />
world, the exception being cask conditioned beer produced primarily in the UK (Chapter<br />
21).The process, also called `lagering' whenreferring tobottom fermented beers, used to<br />
occupy several weeks or even months, but now is often completed in one to two weeks<br />
<strong>and</strong> sometimes in considerably less. It follows that accelerated processes of maturation<br />
have been developed. Traditionally, maturation involves asecondary fermentation <strong>and</strong> is<br />
effected by the small amount of yeast remaining in the beer when it is transferred from<br />
the fermenting vessel. This yeast can utilize fermentable carbohydrates remaining in the<br />
beer at the end of primary fermentation or small quantities of fermentable carbohydrate<br />
added in the form of `priming sugar'. In some systems wort is added to provide the<br />
fermentable material or actively fermenting wort when the process is called `krausening'.<br />
The carbon dioxide that is produced dissolves in the beer because the vessel is closed <strong>and</strong><br />
the beer becomes `conditioned'. Other gases <strong>and</strong> volatile substances are produced during<br />
maturation which are deleterious to beer flavour, e.g., hydrogen sulphide <strong>and</strong> some<br />
diketones. Occasional, but systematic, release of pressure on the maturation vessel will<br />
allow the venting of these compounds to atmosphere.<br />
Green beer is hazy as well as having an unacceptable flavour. During maturation,<br />
clarification of the beer takes place. This is by natural sedimentation in the cold<br />
( 1ëC, 30ëF) of protein <strong>and</strong> polyphenol complexes, but this process can be enhanced<br />
<strong>and</strong> considerably hastened by physical <strong>and</strong> chemical means <strong>and</strong> this is now common<br />
brewery <strong>practice</strong>. Stabilization of the beer is also an important aspect of maturation. The<br />
objective is to ensure that turbidity owing to chemical precipitation or growth of microorganisms<br />
does not occur or, in the case of chemical precipitation, does not recur when<br />
the beer is clear <strong>and</strong> stable. During maturation, treatments can be made to the beer to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
adjust its flavour <strong>and</strong> colour by the use of caramel orother colouring materials <strong>and</strong> by the<br />
use of various post-fermentation hop treatments for both bitterness <strong>and</strong> aroma.<br />
The final treatment for beer before packaging is filtration <strong>and</strong> beers which have been<br />
effectively matured <strong>and</strong> stabilized by whatever means are easier to filter. This is of<br />
importance in large breweries where interruptions in beer filtration are extremely<br />
damaging to overall efficiency. The separation of sediment or `tank bottoms' from the<br />
maturing beer either by, or before filtration is, therefore, crucial to success. The disposal<br />
of this sedimented material is becoming aproblem in some areas, e.g., the UK.<br />
Beers brewed at higher than sales gravity are diluted before packaging <strong>and</strong> this often<br />
occurs after maturation but under the same departmental control. The department in the<br />
brewery in which maturation <strong>and</strong> clarification is carried out is often referred to as the<br />
process department. Beer changes or treatments after primary fermentation but before<br />
packaging therefore comprise:<br />
· maturation; flavour <strong>and</strong> aroma changes<br />
· stabilization against non-biological haze<br />
· carbonation<br />
· biological stabilization (pasteurization or sterile filtration, Chapter 21)<br />
· clarification <strong>and</strong> filtration.<br />
Certain special beer types can also derive from further treatments post fermentation.<br />
These include low-alcohol <strong>and</strong> non-alcoholic beers <strong>and</strong> so-called ice beers.<br />
15.2 Maturation: flavour <strong>and</strong> aroma changes<br />
The flavour changes that take place as abeer matures are profoundly important in<br />
developing the character <strong>and</strong> hence the br<strong>and</strong> identity of the beer. Successful br<strong>and</strong>s<br />
generally have stable flavours <strong>and</strong> so can be recognized by consumers. This applies to<br />
national <strong>and</strong> international br<strong>and</strong>s but even local br<strong>and</strong>s will be expected to display<br />
consistent taste. Flavour improves during the maturation process but this flavour<br />
improvement is difficult to characterize <strong>and</strong> optimize. There is the added factor of the<br />
effect of oxygen, which will generally cause adverse flavour changes, <strong>and</strong> so any<br />
discussion of flavour maturation must include astudy of ways of preventing oxidation.<br />
Maturation is carried out in many different ways <strong>and</strong> it is difficult, therefore, to establish<br />
the underlying principles. What is clear is that processes of secondary fermentation <strong>and</strong><br />
subsequent cold storage are involved that frequently took several months but now are<br />
usually completed in around two weeks or less, so investment in the fixed assets of<br />
storage tanks has been reduced.<br />
15.2.1 Principles of secondary fermentation<br />
Secondary fermentation permits continued activity by the yeast at areduced rate limited<br />
by the low temperature <strong>and</strong> the lower yeast count in the beer. Traditionally after primary<br />
fermentation (Chapter 12) the beer would pass to the conditioning or maturation vessel<br />
<strong>and</strong> would contain 1±4 million yeast cells/ml of beer <strong>and</strong> about 4ë of gravity (1.1%<br />
fermentable extract). There are many temperature regimes which are subsequently<br />
applied, <strong>and</strong> they represent compromises between promoting production of carbon<br />
dioxide <strong>and</strong> hence providing condition to the beer <strong>and</strong> allowing the removal of<br />
undesirable flavour compounds. The beer was cooled, traditionally to 8 ëC (46 ëF) at the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
end of primary fermentation to remove most of the surplus yeast before transfer to the<br />
warm maturation vessel. In this process the remaining yeast becomes re-suspended <strong>and</strong><br />
there is asmall uptake of oxygen, which activates the yeast to start the slow secondary<br />
fermentation. This results in the conversion of many unwanted flavour compounds into<br />
flavourless products (O'Rourke, 2000).<br />
Flocculentyeastsseparateeasilyattheendofprimaryfermentation<strong>and</strong>conditionscan<br />
be adjusted such that sufficient yeast can be retained in the beer to effect the flavour<br />
changes required in maturation (Chapter 12). Powdery yeasts, not separating effectively,<br />
may ferment too fully in secondary fermentation <strong>and</strong> remove all residual extract <strong>and</strong> may<br />
remain in suspension making clarification difficult. These different situations provide<br />
constant challenges to fermentation <strong>and</strong> maturation management. In any event the yeast<br />
must have access to fermentable carbohydrate for the process to succeed. This<br />
carbohydrate is provided, as above, by residual gravity in the beer or by the addition of<br />
sugar by priming or by krausening. Krausening is the addition of wort from the active<br />
`krausen' stage of the primary fermentation usually at 5±10% by volume of the green<br />
beer. In shorter secondary fermentation regimes yeast activity must be intense to achieve<br />
carbonation, purging of the undesirable volatiles, removal of all residual oxygen <strong>and</strong><br />
chemical reduction of many compounds. This leads to immediate improvement of flavour<br />
<strong>and</strong> aroma <strong>and</strong> flavour stability.<br />
15.2.2 Important flavour changes<br />
Several important groups of compounds have been identified as changing during the<br />
maturation of beer with consequent positive effect on beer flavour. The most important<br />
are: diketones (especially diacetyl), sulphur compounds, aldehydes, <strong>and</strong> volatile fatty<br />
acids.<br />
Diketones<br />
Diacetyl <strong>and</strong> 2,3-pentanedione are produced in all brewery fermentations (Chapter 12).<br />
Diacetyl in particular has an intense sweet, butterscotch flavour. This cannot be tolerated<br />
in lager beers <strong>and</strong> the concentration in finished beer should be < 0.1 mg/l. A period of<br />
warm conditioning (2±3 days at 14±16 ëC, 59 ëF) is very effective in reducing the diacetyl<br />
content of beer. The precursors of diacetyl <strong>and</strong> 2, 3-pentanedione, -acetolactate <strong>and</strong> -<br />
acetohydroxybutyrate, are excreted by the yeast <strong>and</strong> are non-enzymically converted in the<br />
green beer to the vicinal diketones by oxidative decarboxylation. The level of the<br />
acetohydroxy acids in beer is a function of the yeast strain <strong>and</strong> is enhanced by conditions<br />
of rapid yeast growth. Yeast cells will not assimilate exogenous acetohydroxy acids but<br />
will readily take up <strong>and</strong> reduce diacetyl <strong>and</strong> 2, 3-pentanedione to acetoin <strong>and</strong> 2, 3-pentane<br />
diol, which have no adverse flavours. The rate of this reaction is dependent on the yeast<br />
strain, how it has been stored, <strong>and</strong> its age. This forms the basis of effective diacetyl<br />
removal from green beer; the yeast must be in a healthy metabolic state to carry out the<br />
reduction efficiently. Frequent causes of inability to control the concentration of vicinal<br />
diketones in beer are yeast of the wrong strain or yeast in poor health perhaps accelerated<br />
by the too rapid onset of fermenter cooling, causing the yeast to separate. If active yeast is<br />
not present diacetyl will not be reduced.<br />
The effective <strong>and</strong> reproducible removal of vicinal diketones from beer is important in<br />
overall brewery efficiency. Delays in diacetyl removal result in delays in filtration <strong>and</strong><br />
hence delays in the supply of beer to packaging. This has stimulated work to examine<br />
ways of accelerating the removal of diketones. The most effective way remains the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
choice of an appropriate strain of yeast. However, in some cases strains are used which<br />
have other desirable properties but are inherently poor at diketone removal. The use of a<br />
commercially available -acetolactate decarboxylase enzyme has been proposed<br />
(Hanneman, 1999). It was shown, in tube fermentations <strong>and</strong> full-scale brewery trials,<br />
that maturation was accelerated <strong>and</strong> the need for cold beer lagering eliminated orreduced<br />
to an absolute minimum. The optimal dose of enzyme depends on the yeast strain <strong>and</strong><br />
wort composition. The enzyme is derived from Bacillus subtilis, which contains agene<br />
from Bacillus brevis. The enzyme is now used in many parts of the world (Jepsen, 1991).<br />
It is not approved for use in the UK.<br />
Anovel method of maturation, allowing very rapid reduction of the diacetyl content, has<br />
been described using immobilized yeast in abioreactor (Pajunen et al., 1991; Pajunen <strong>and</strong><br />
JaÈaÈskelaÈinen, 1993, see Chapters 13 <strong>and</strong> 14). Yeast cells are immobilized on beads of<br />
DEAE-cellulose (diethylaminoethylcellulose). The brewing yeast is first separated from the<br />
green beer by centrifugation to give avery low yeast count. It is then heated to 90ëC<br />
(195ëF) for 8±10 minutes to convert all the -acetolactate to diacetyl. The beer is then<br />
passed through the immobilized yeast reactors <strong>and</strong> the diacetyl is converted to acetoin. The<br />
beer is then ready for filtration <strong>and</strong> packaging. The technique has been used commercially<br />
(Pajunen <strong>and</strong> JaÈaÈskelaÈinen, 1993) to operate at an annual volume of one million hl. In this<br />
system there are four bioreactors with amaximum total flow of 14m 3 /h, which corresponds<br />
to the centrifuge capacity <strong>and</strong> the time needed for rapid emptying of primary fermenters. In<br />
this system the maturation process is reduced to a matter of hours (frequently
There has been some discussion about the purging effect on volatiles of the carbon<br />
dioxide produced during beer maturation. There is some evidence that hydrogen sulphide<br />
is more effectively purged from beers where carbon dioxide production is vigorous<br />
(Zangr<strong>and</strong>o <strong>and</strong> Girini, 1969).<br />
Aldehydes<br />
Acetaldehyde in particular can affect beer flavour. This arises from the oxidation of<br />
ethanol <strong>and</strong> can occur if transfer of the beer from primary fermentation to maturation is<br />
carelessly carried out giving the opportunity for oxygen uptake by the beer. During<br />
normal maturation the acetaldehyde concentration will decrease to 2±7 mg/l. Acetaldehyde<br />
can be detected at about 10 mg/l in a lightly flavoured pilsen-type beer, when it<br />
gives a flavour of green apples. It is less easily detected in ales <strong>and</strong> is a characteristic of<br />
some ale flavours. Concentrations of > 35 mg/l should be avoided.<br />
Volatile fatty acids<br />
Beer conditioning temperature is very important in determining the excretion of C4 to C10<br />
fatty acids. Synthesis of short chain fatty acids by yeast stops at the onset of maturation.<br />
C8 fatty acid increases in concentration during fermentation <strong>and</strong> this is replaced in<br />
maturation by C10 acid. Glycerides <strong>and</strong> phospholipids are synthesized during maturation<br />
<strong>and</strong> so there is a general trend for a reduction in volatile acids as ageing proceeds. This<br />
trend can be reversed if maturation is extended too far. There can then be a rise in the<br />
concentration of free fatty acids owing to the hydrolysis of reserve glycerides, with<br />
consequent adverse effects on flavour. If a high maturation temperature is maintained for<br />
too long then there can be a slow excretion of C10 acid (capric acid), which has a flavour<br />
threshold of 10 mg/l <strong>and</strong> this is undesirable. Maturation is seldom controlled specifically<br />
from the viewpoint of controlling volatile fatty acids. Provided that the warm<br />
conditioning is not extended beyond the time needed to reduce the concentration of<br />
vicinal diketones, <strong>and</strong> cold storage is controlled to the time required for stabilization of<br />
haze precursors, there should not be a problem.<br />
15.2.3 Techniques of maturation<br />
Technique varies widely from brewery to brewery. In general, for bottom fermented beers<br />
the technique must ensure the production of a balanced beer flavour with the minimum<br />
concentration of diacetyl. For brewery conditioned top fermented beers, the technique<br />
centres around creating the required haze stability as the more robust flavour of the top<br />
fermented ale derives from primary fermentation <strong>and</strong> diacetyl reduction is not a problem.<br />
Lager methods<br />
Beer is transferred to a separate tank for maturation in traditional lagering methods.<br />
Before the end of primary fermentation cooling is applied to the cone of the fermenter to<br />
achieve a temperature of 5 ëC (41 ëF). The remainder of the beer above the cone (at least<br />
95% of the volume) is at a higher temperature to ensure effective diacetyl removal.<br />
Cooling below 5 ëC (41 ëF) is not necessary <strong>and</strong> runs the risk of cooling the beer beyond<br />
its point of maximum density when inversion of flow around the fermenter may occur<br />
(Andrews, 1997). About 24 hours after applying cooling an initial removal of yeast is<br />
usually carried out. This yeast is usually discarded. When diacetyl reduction is complete<br />
the remainder of the beer is slowly cooled to 5 ëC (41 ëF) to complete the maturation by<br />
adjusting the flavour volatiles. A sudden fall in temperature must be avoided or the shock<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
may induce the yeast to excrete protease enzymes that could be detrimental to foam<br />
stability. This cooling may take from two to nine days. At the end of this period asecond<br />
removal of yeast is usually made. The beer is now completely cooled to at least 1ëC<br />
(30ëF). The period of colloidal stabilization now takes place at 1to 2ëC (30 to 28ëF)<br />
for two to three days. Afinal yeast removal is made before filtration. The colloidal<br />
stabilization temperature must be maintained throughout the whole vessel.<br />
There are many variations on this technique which would still be regarded as<br />
traditional. Indeed the times can be much longer (Miedaner, 1978) with storage times at<br />
1ëC (30ëF) being up to six weeks. In this situation up to 40% of the capital cost of the<br />
brewery could be in the tanks required to condition beer (Coors, 1977). Fermentation <strong>and</strong><br />
maturation can alsobecarriedout inonevessel soavoiding transfer betweenprimary<strong>and</strong><br />
secondary fermentation.<br />
Because of the high capital costs associated with building tanks <strong>and</strong> the working<br />
capital of stored beer, brewers have sought to reduce maturation times. Yeasts with the<br />
ability to reduce diacetyl rapidly have been used <strong>and</strong> warm storage times have been<br />
almost eliminated. There are almost as many variations on maturation technique as there<br />
are breweries. At least one successful system comprises: fermenter filling 20 hours,<br />
primary fermentation 72 hours, warm storage 48 hours, cooling 48 hours, cold storage at<br />
1ëC (30ëF) 36 to 48 hours. Total time is less than ten days. These systems can produce<br />
sound beer. However things can go wrong <strong>and</strong> they are susceptible to variations in raw<br />
material quality. To escape from periods of uncertainty over maturation <strong>and</strong> hence<br />
flavour stability, brewers need to keep in mind the traditional principles of the process.<br />
One of the most successful systems for reducing maturation time seems to be the system<br />
of using immobilized yeast in abioreactor (Section 15.2.2).<br />
Ale methods<br />
The traditional method for maturation <strong>and</strong> conditioning of ale is cask conditioning<br />
(Chapter 23). Now many types of ale are brewery conditioned <strong>and</strong> filtered <strong>and</strong> sold in<br />
kegs (Chapter 23). Ale fermentations are rapid <strong>and</strong> vigorous <strong>and</strong> usually completed in 48<br />
to 60 hours (Chapter 12) at temperatures of up to 24ëC (75ëF) before being cooled to<br />
Turbulence causes the loss of carbon dioxide <strong>and</strong> the uptake of oxygen. Beer must<br />
therefore be moved gently through correctly sized pipes.<br />
Movement of beer from one tank to another, e.g., from fermenter to maturation vessel<br />
is often encouraged by applying top pressure of carbon dioxide to the donor tank after<br />
flushing the mains <strong>and</strong> the receiving tank with the gas. But carbon dioxide is expensive<br />
<strong>and</strong> the cheaper nitrogen gas is now often used. The high density of carbon dioxide,<br />
relative to air, can be exploited. Athin blanket of carbon dioxide can be produced above<br />
the beer in the donor tank, <strong>and</strong> atop pressure of air can be used to move the beer out of<br />
the tank. Asmall amount of carbon dioxide can then be injected into the beer en route to<br />
the receiving tank where the gas escapes to provide the blanket of carbon dioxide above<br />
the beer now in the receiving tank. As the tank fills the air within it is displaced upwards<br />
<strong>and</strong> does not come into contact with the beer. Alternatively the receiving tank can be<br />
filled with deaerated liquor (containing
Caramels used in brewing are electropositive <strong>and</strong> can have acolour of 50,000ëEBC<br />
<strong>and</strong> contain 65% solids. In Europe, caramels are classified <strong>and</strong> those used in brewing are<br />
designated E150c caramels. These caramels can contain 4-methyl imidazole which is<br />
toxic to rabbits, mice <strong>and</strong> chickens (Hasimoto, 1973) <strong>and</strong> more recent research has<br />
revealed the presence of 2-acetyl-4-tetrahydroxybutyl imidazole (THAI), which has been<br />
found to react as a anti-pyridoxine factor in rats leading to lympholeucocytopaenia<br />
(Leach, 1989). For these reasons caramels used in brewing must have THAI levels<br />
fermentation <strong>and</strong> so change the aroma of beer. Some of these preparations contain resins<br />
as well as oils <strong>and</strong> are known as oil-rich extracts. They must be added to beer before<br />
filtration to avoid the possibility of haze developing in the beer during storage.<br />
Recently (Marriott, 1999) there have been further developments to produce aroma<br />
products that are completely soluble in bright beer (see also Chapters 7<strong>and</strong> 8). These can<br />
give potentially 100% utilization <strong>and</strong> so acompletely reproducible effect on aroma.<br />
Essences have been produced that can reproduce the effect of late copper hopping <strong>and</strong><br />
even the dry hopping of cask ale (Chapter 7). Essentially the insoluble hydrocarbon<br />
fraction of the oil must be removed to produce asesquiterpeneless oil. This oil can then<br />
be further fractionated to produce individual products with `spicy' or `floral' characters.<br />
These essences can be added to bright beer at rates of 50 to 100 g/l <strong>and</strong> have profound<br />
effects on aroma <strong>and</strong> flavour. They are normally supplied as 1% solutions in ethanol.<br />
Blending<br />
Beers can be blended post fermentation <strong>and</strong> this yields more uniform products <strong>and</strong><br />
provides further opportunities to deliver avolume of beer to its final specification.<br />
Recovered beer can also be added post fermentation but this is atechnique requiring<br />
considerable care, often dem<strong>and</strong>ing excessive pasteurization of the beer with a<br />
consequent adverse effect on flavour. For this reason recovered beer is best added<br />
before fermentation, e.g., to the whirlpool (Section 15.5).<br />
Sulphur dioxide<br />
Sulphur dioxide is both anatural product of fermentation (Chapter 12) <strong>and</strong> can be added<br />
to beer after fermentation. It provides a measure of protection against flavour<br />
deterioration by oxidation <strong>and</strong> has bacteriostatic properties. Maximum levels of sulphur<br />
dioxide in beer are usually governed by statute <strong>and</strong> vary in different countries. In the<br />
European Union the maximum permitted level is 20mg/l (as SO2), except for cask<br />
conditionedbeerwhenthelevelis50mg/l.Sulphurdioxide isusuallyaddedassodiumor<br />
potassium metabisulphite <strong>and</strong> can so be added along with finings or priming sugar.<br />
Water<br />
It is an accepted strategy in many breweries to brew at ahigher gravity than that at which<br />
the beer is subsequently sold (Chapter 6). Many studies have been made on the effect of<br />
gravity (wort strength) on the properties of the resultant beer <strong>and</strong> the physiological health<br />
of the yeast after subsequent generations at high gravity (>1060 or 15ëP) (Pfisterer <strong>and</strong><br />
Stewart, 1975; Stewart et al., 1999). Dilution of the fermented high gravity beer is best<br />
effected after beer filtration <strong>and</strong> the quality of the diluting water used is of the utmost<br />
importance (Chapter 3). This needs to be of brewing quality because this water is to be<br />
drunk by the consumer of the product. It must be free from taint, sterile <strong>and</strong> deaerated.<br />
Thisisnormallyachievedbytreatingthewaterbypassingitthroughapurpose-builtplant<br />
involving atrap filter to remove solids, acarbon filter to remove chlorine residues <strong>and</strong><br />
inert gas stripping or vacuum stripping to remove oxygen. Sterilization is often effected<br />
by the use of uv light or by filtration through asterilizing sheet filter (0.45 m). The<br />
oxygen level in the water should be
the beer stream following accurate measurement of density often by the oscillating Utube<br />
method.<br />
15.2.5 Maturation vessels<br />
Vessels for the maturation of ales or lagers are normally cylindrical in shape. They are<br />
similar to fermentation vessels (Chapter 14). They can be horizontal or vertical in aspect.<br />
Tanks can be made of stainless steel, mild steel with aglass, enamel or plastic epoxy<br />
lining, or more rarely of aluminium. Horizontal tanks are usually of 100 to 500hl (60±<br />
300 imp. brl) capacity, but the vertical cylindroconical tanks can be up to 6,500hl (4,000<br />
imp. brl). Horizontal tanks are frequently built inside the brewery in warm or cold<br />
conditioning rooms. Vertical tanks are externally clad <strong>and</strong> usually st<strong>and</strong> in the open.<br />
Tanks are normally fitted with impellers for mixing <strong>and</strong> with atemperature control<br />
system.Temperatureisusuallycontrolledbyacoolingjacketsuppliedwithbrine,ethanol<br />
as industrial methylated spirit (IMS), or propylene glycol. Ethylene glycol has been used<br />
but it is highly toxic. Older designs incorporated internal cooling coils supplied with<br />
chilled water or IMS. Frequent inspection of the integrity of the coils is essential if a<br />
coolant other than water is used, as small leaks would be disastrous for beer quality.<br />
The escape of carbon dioxide gas is limited by closing the vessel <strong>and</strong> hence pressure<br />
regulation is required. Devices are used to control gas pressure up to 1.4 bar (20lb./in. 2 ).<br />
Water column or mercury column manometers were once used but now weight loaded<br />
valves are employed, which can be set to open <strong>and</strong> release pressure at adefined level<br />
(Fig. 14.3).<br />
Materials of construction <strong>and</strong> vessel size<br />
Construction is similar to that of fermenting vessels (Section 14.3.2). The preferred metal<br />
of construction would now be type AISI 304 stainless steel, it is rare that conditions are<br />
such that the much more expensive AISI 316 is needed. The inner surface of the tank<br />
shouldbeassmoothaspossiblesoastoprovidenosurfaceindentationsforthelodgingof<br />
soil (Section 14.3.2).<br />
Size is important when the vessels are to be used for fermentation <strong>and</strong> maturation, i.e.,<br />
uni-tanks (Fig. 15.1). This relates to the hydrostatic pressure on the yeast during<br />
fermentation <strong>and</strong> agenerally acceptable height for afermenter is now not greater than<br />
15m(about 50ft.). In apurpose-built maturation tank there is no such restriction <strong>and</strong><br />
tanks up to 30m(almost 100ft.) <strong>and</strong> even 40m(130ft.) in height have been built. Size<br />
usually relates to brewery throughput <strong>and</strong> brew lengths. Arule of thumb is that the size<br />
should be equivalent to ahalf-day production, larger tanks take too long to fill <strong>and</strong> will<br />
contain beers of variable age <strong>and</strong> hence potentially variable final flavour. Tank diameters<br />
are usually 3.50±4.75m(11.5 to 16ft.) <strong>and</strong> the cone angle is 60ë to 75ë. The ratio of<br />
diameter to beer height in the cylinder can vary from 1:1 to 1:5. The volume of the<br />
headspace in the tank relates to the work to be done. Avessel used solely for cold<br />
maturation will require aheadspace volume of 5% of the total, whereas if diacetyl<br />
removal in the warm is included the volume excess should be 10%. Afully equipped<br />
fermenting vessel will need 25% headspace.<br />
Cooling<br />
The details of temperature control in fermentation were discussed in Chapter 14 <strong>and</strong><br />
many of the principles apply to the maturation vessel. Chilled water is unsuitable for use<br />
as a coolant below 2 ëC. Propylene glycol has the disadvantage of being extremely<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
TPA<br />
SB<br />
PT<br />
Cooling<br />
panel<br />
Cooling<br />
panel<br />
Cooling<br />
panel<br />
TP<br />
Gross volume<br />
Working<br />
volume<br />
viscous (13.4mPa at 0ëC). Ethanol (IMS) is often the preferred coolant, but it should be<br />
noted that both IMS <strong>and</strong> propylene glycol will oxidize slowly to acid products which will<br />
attack mild steel pipework <strong>and</strong> cause sludge deposits. Exposure to air should therefore be<br />
minimized <strong>and</strong> hence solution strengths should be maintained (30% for ethanol <strong>and</strong> 40%<br />
for propylene glycol). Sometimes acorrosion inhibitor is included with the secondary<br />
coolant. There has been renewed interest in direct expansion systems using aprimary<br />
refrigerant such as ammonia. Ammonia has no ozone depleting effect <strong>and</strong> therefore has<br />
this major advantage over the fluorinated hydrocarbons, which are now being withdrawn<br />
as refrigerants. However ammonia is avery corrosive, dangerous <strong>and</strong> pungent gas <strong>and</strong><br />
brewers still tend to favour indirect systems (Section 14.3.3).<br />
Several designs of cooling jacket are available (Chapter 14). The requirements are<br />
more simple than in the fermenter or combined vessel <strong>and</strong> the choice often depends on<br />
the manufacturer's capability. Coils can be wound onto the vessel section <strong>and</strong> seam<br />
welded (limpet coils). An alternative uses aprofile plate, which is preformed <strong>and</strong> spotwelded<br />
on tothe vessel,avariation ofthe preformed plate is the `dimple jacket', which is<br />
also preformed, <strong>and</strong> spot welded (Fig. 15.2). An inflated `quilted jacket' is afurther<br />
alternative most frequently used in direct expansion systems with ammonia. Most vessels<br />
will have two cooling sections on the cylinder of the vessel <strong>and</strong> adimple jacket on the<br />
cone. The cone section cooler will usually be used at the end of primary fermentation to<br />
aid yeast flocculation.<br />
TP<br />
Bottom manway<br />
Fig. 15.1 Schematic representation of dual-purpose fermenting vessel (1600hl, 1000 imp. brl); TP,<br />
temperature probes; PT, pressure transducer; TPA, top plate assembly containing a pressure relief<br />
valve (0.8 bar, 12 lb./in. 2 ), anti-vacuum valve (55 mm WG), CO 2/CIP combination valve, top pressure<br />
transmitter, access hatch, inspection light, trace heating <strong>and</strong> cabin for weather protection; SB, spray<br />
ball operating at 5 bar (74 lb./in. 2 ), 80 ëC (175 ëF), 17 m 3 /h; Cooling panels contain coolant at 4.5 bar<br />
(66 lb./in. 2 ); Height 16.0 m; Diameter 4.25 m; Working temperature, 1 to 90 ëC (30±190 ëF); Working<br />
pressure, 0.7 bar (10 lb./in. 2 ); Wort depth 13.5 m; Weight empty 14 t; Weight full 212 t; Gross volume<br />
1922 hl; Working volume, 1640 hl; Freeboard volume 282 hl (17% of working volume); Cone volume<br />
147 hl (9% of working volume; included cone angle 70ë) (Barnes, 2001).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
TP<br />
10.5 m<br />
3.0 m 12.0 m<br />
1.0 m
Plain dimple Limpet<br />
Pressed<br />
corrugated<br />
Inflated<br />
panel<br />
Fig. 15.2 Designs ofcoolingjackets fordual-purpose fermenting vessels: plainjacket, notusually<br />
used because it requires thick vessel construction to resist external pressures; limpet, provides a<br />
flow-path for coolant, but again relatively heavy; pressed corrugated, lighter construction than<br />
limpet <strong>and</strong> widely used for secondary refrigerant systems; inflated panel, low volume <strong>and</strong> suited to<br />
primary refrigerant systems acting as an evaporator (Barnes, 2001).<br />
Refrigerationloadinabreweryisaconsiderablefixedcost<strong>and</strong>thereforethedesignofthe<br />
controlsystemtoensurethemostefficientuseofenergyisimportant.Thenumberofpumps<br />
for secondary coolant should be minimized in the design. Avariable speed drive in the<br />
secondary coolant system will maintain aconstant pressure differential in the supply <strong>and</strong><br />
will help to minimize losses. There should also be the facilityto shut downthe system if no<br />
vesselsrequirecooling.Thepositionofthetemperaturesensoriscriticaltoefficientcooling.<br />
Thesensorshouldbepositionedatbetweenone-third<strong>and</strong>one-halfofthevesselheightfrom<br />
thebottom<strong>and</strong>shouldprotrudeabout500mm(20in.)intothetanktoavoidanytemperature<br />
effectfromtheinternalsurfaceofthetank.Informationfromthesensorinamodernbrewery<br />
will be fed back to acomputer at which the desired temperature regimes may be set.<br />
Cleaning-in-place (CIP)<br />
Cylindroconical maturation vessels are cleaned by CIP systems. The principle of this<br />
cleaning is very similar to that described for fermenting vessels (Section 14.3.3) <strong>and</strong><br />
discussed in overall detail in Chapter 17. Maturation vessels do not generally have such a<br />
high organic soil level (yeast deposits) as fermenting vessels <strong>and</strong> therefore respond to<br />
cleaning with acids, which is usually cheaper than caustic alkali based systems using a<br />
sequesterant. However, as with fermenters to save costs of detergent, the bulk of the soil<br />
should be removed with jets of water in an impact system. The jets used vary enormously<br />
in design but are often similar to those used for cleaning fermenting vessels, where the<br />
nozzles on the head rotate in one plane whilst the head itself rotates at right-angles on the<br />
support pipe (Fig. 14.6). The basic sequence of CIP is similar to that used for fermenters<br />
(Section 14.3.3) with slightly shorter time sequences.<br />
Insulation<br />
Similar principles apply to those discussed in Chapter 14 in relation to fermenting<br />
vessels. Horizontal tanks are usually contained within aroom in the brewery which is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
temperature controlled. The tanks themselves are not insulated. Vertical tanks require<br />
insulation. Insulation materials often contain chloride ions <strong>and</strong> a chloride barrier is<br />
needed against the stainless steel of the vessel to prevent attack of the surface by chloride<br />
(Section 14.3.2).<br />
15.3 Stabilization against non-biological haze<br />
Competition between brewers is intense <strong>and</strong> the quality <strong>and</strong> consistency of their beers is<br />
paramount. This dem<strong>and</strong>s that the beers following maturation should not only have<br />
desirable, stable flavours but must also display stability with respect to haze, i.e., the<br />
beers must be bright <strong>and</strong> remain so during the period from dispatch from the brewery to<br />
drinking. Therefore, in addition to removing yeast, beers must have the precursor<br />
constituents of haze removed to ensure long-term stability. Beer haze <strong>and</strong> its chemistry<br />
are considered in Chapter 19. In this Chapter we focus on the removal of haze-forming<br />
materials to ensure the production of astable beer.<br />
Arange of substances can cause non-biological haze in beer:<br />
· -glucans, which can often lead to hazes not easily seen by eye but which cause high<br />
levels of light scattering in 90ë haze meters<br />
· -glucans (starch), which can behave similarly to -glucans<br />
· pentosans, which may be derived from wheat based adjuncts<br />
· dead bacteria from malt<br />
· oxalate from calcium deficient worts.<br />
However, the most common, important <strong>and</strong> troublesome type of non-biological haze is<br />
that deriving from the cross-linking of proteins <strong>and</strong> polyphenols <strong>and</strong> it is the elimination<br />
of the precursors of these polymers to which beer stabilization treatments are directed.<br />
The most effective beer treatment with respect to haze stability is the cold storage of<br />
the beer for about seven days at 1 to 2ëC (30±28ëF). This technique allows a<br />
reduction in the cost of other beer treatments designed to remove potential haze-forming<br />
proteins <strong>and</strong> polyphenols. However, brewers frequently wish to accelerate the process of<br />
haze stabilization <strong>and</strong> achieve greater stability than is possible with cold storage alone.<br />
15.3.1 Mechanisms for haze formation<br />
Colloidal haze in beer arises from the formation of protein-polyphenol complexes during<br />
beer storage (Gopal <strong>and</strong> Rehmanji, 2000, <strong>and</strong> Chapter 19). Fresh beer contains acidic<br />
proteins<strong>and</strong> numerous polyphenols. These can come together byloosehydrogen bonding<br />
but theassociations formed aretoo small tobeseen bythe nakedeye. Thesepolyphenols,<br />
called flavanoids, can further polymerize <strong>and</strong> oxidize to produce condensed polyphenols,<br />
which have been called tannoids (Chapon, 1994). These tannoids can `bridge' by<br />
hydrogen bonding across anumber of proteins to form areversible chill haze (Fig. 15.3).<br />
This haze forms at around 0 ëC (32 ëF) but redissolves when the beer is warmed to 15 ëC<br />
(59 ëF). After further storage of the beer strong bonds can form between the tannoids <strong>and</strong><br />
proteins <strong>and</strong> irreversible, permanent haze is formed. The rate at which this haze is formed<br />
<strong>and</strong> its extent of formation depends on the raw materials used in wort preparation <strong>and</strong> the<br />
process conditions. This `model' suggests that effective stabilization should be achieved<br />
by removing from the beer the constituents of the haze, i.e., the `tannin sensitive' proteins<br />
<strong>and</strong>/or the polyphenols.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Simple<br />
flavanoids<br />
Protein<br />
Oxidized flavanoids<br />
= tannoids<br />
Proteins<br />
Tannoids<br />
Proteins<br />
Chill haze<br />
Permanent haze<br />
Fig. 15.3 Models of chill <strong>and</strong> permanent hazes development in beer (Gopal <strong>and</strong> Rehmanji, 2000).<br />
An alternative model of haze-formation has been proposed (Siebert et al., 1996 <strong>and</strong><br />
Chapter 19). This suggests there are afixed number of binding sites on haze-forming<br />
proteins (proline residues) <strong>and</strong> that haze-forming polyphenols have two binding sites,<br />
through which they can jointo two adjacent protein molecules. If there is an excess of<br />
protein over polyphenol then the polyphenol is involved in binding just two protein<br />
molecules together <strong>and</strong> these dimers do not constitute insoluble complexes. If the<br />
amount of polyphenol greatly exceeds that of protein then there is a shortage of protein<br />
binding sites <strong>and</strong> again haze complexes will not be formed. Hazes are therefore formed<br />
when there are equivalent amounts of protein <strong>and</strong> polyphenol in the beer. This model<br />
suggests an alternative strategy for the prevention of haze, i.e., substantially increase the<br />
amount of either protein or polyphenol. This is not a favoured approach <strong>and</strong> most<br />
brewers will seek to reduce levels of either the proteins or polyphenols or most likely<br />
both.<br />
15.3.2 Removal of protein<br />
All of the haze-forming protein in beer comes from malt. Proteins which are particularly<br />
liable to cause haze are rich in proline <strong>and</strong> have molecular weights > 10,000. A<br />
concentration of 2 mg/l will give a haze value in beer of > 1.0 EBC formazin units<br />
(Chapter 19) which will give a perceptible turbidity to the beer. However, other proteins<br />
can also form haze <strong>and</strong> some of these have good foam potential. It is not, therefore, a<br />
simple matter to categorize the proteins responsible for haze-formation in beer. It can<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
simply be concluded that the presence of hydrophobic groups on the surface of protein<br />
tertiary structure increases the capacity to form haze <strong>and</strong> the capacity to improve foam.<br />
Lowering the overall protein concentration will increase haze shelf-life. The use of low<br />
nitrogen malts (TN 40%) have inherently poor<br />
foam potential. It is thus theremoval of protein thatisimportant. Proteins can essentially be<br />
removed by hydrolysis; usually by enzymes, by precipitation or by adsorption.<br />
Hydrolysis<br />
The usual enzyme to use is papain derived from papaya fruit although bromelin from<br />
pineapple <strong>and</strong> an acidic protease from Bacillus subtilis have been used (Chapter 2). The<br />
optimum dose is 5±15 10 6 units/hl (1 unit is the amount of enzyme generating 1 g<br />
soluble tyrosine/hour under rigidly st<strong>and</strong>ardized test conditions). This usually translates to<br />
0.5±4.0g/hl depending on the formulation. The enzyme needs acontact time <strong>and</strong> is usually<br />
added to the beer during maturation. Some preparations can be added to final beer but they<br />
must be sterile <strong>and</strong> totally soluble. This treatment is cheap but is risky. Proteolytic enzymes<br />
will damage beer foam proteins <strong>and</strong> some can survive pasteurization so will continue to<br />
have an effect on foam during beer storage. There can also be an adverse effect on flavour<br />
stability by the liberation of free thiol groups from peptides in the beer.<br />
Precipitation<br />
Since proteins form haze complexes with polyphenols it is not surprising that aclass of<br />
polyphenol can be used for the removal of haze sensitive protein. Thus anionic tannic acid<br />
reacts with cationic proteins to form an insoluble complex. This is as aresult of interaction<br />
between ketone groups on the tannin <strong>and</strong> nucleophilic groups (SH, NH 2)on the protein. The<br />
most effective tannins are known as gallotannins extracted from gall nuts or from the sumach<br />
tree.Thetanninpreparationwouldnormallybeaddedenroutetomaturationat0ëC(32ëF)at<br />
5±9g/hl<strong>and</strong>wouldrequireatleast24hourscontacttimeforfulleffectiveness.Essentiallythe<br />
tannins act as aprecipitant of the tannin sensitive proteins (Anderson et al., 2000) <strong>and</strong> when<br />
used in this way in the cold tank alarge volume of `tank bottoms' is formed. This requires<br />
careful movement of the beer from above the sediment or the use of acentrifuge or filter to<br />
separate beer from the tank bottoms (Section 15.5). If this is not done losses of beer will be<br />
high. For this reason the use of tannic acid for protein removal has generally gone out of favour<br />
in recent years. However, improved preparations of gallotannins have become available<br />
(Musche <strong>and</strong> de Pauwe, 1999). These high molecular weight gallotannins can be dosed into<br />
the beer before filtration into the inlet buffer tank to the filter at 0 ëC (32 ëF) at rates of 2±4 g/hl.<br />
A contact time of 5±25 minutes is claimed to be sufficient. This method of treatment avoids<br />
the production of tank bottoms but is limited by the relatively short filtration run achieved <strong>and</strong><br />
for this reason centrifugation is still often used before the filter. Gallotannins have the added<br />
property of removing metals from beer (Musche <strong>and</strong> de Pauwe, 1999). Iron, aluminium, lead<br />
<strong>and</strong> copper can all be removed by gallotannin treatment <strong>and</strong> filtration. The metals can be<br />
lowered in concentration to below the levels at which the properties of beer are harmed.<br />
Adsorption<br />
Silica gels are used as protein adsorbents. They are produced for brewers in two main<br />
commercial forms: hydrogel (> 30% water) <strong>and</strong> xerogel (< 10% water). To produce a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
silica gel, sulphuric acid <strong>and</strong> sodium silicate solution are mixed under controlled<br />
conditions to produce a `hydrosol' of colloidal polysilicic acid. These colloidal species<br />
can be condensed to produce a hydrogel in which the water phase is immobilized in the<br />
silica matrix (McKeown <strong>and</strong> Earl, 2000). The hydrogel can then be further processed <strong>and</strong><br />
then milled to produce a material of defined particle <strong>and</strong> pore size <strong>and</strong> surface area.<br />
Optimum pore size is in the range 30±120 m diameter with a surface area of around<br />
800 m 2 /g.<br />
Alternatively, the hydrogel can be dried before milling to produce a xerogel where the<br />
surface area can range from 300±800 m 2 /g with again pore sizes in the range 30±120 m.<br />
Protein adsorption takes place on so-called silonol (SiOH) sites on the silica gel <strong>and</strong><br />
adsorption capacity is a function of the number of these sites. Generally, though, the<br />
number of sites exceeds the quantity of available protein.<br />
Silica gels can be added to beer in maturation but their most attractive characteristic is<br />
their usefulness when dosed into the beer stream prior to filtration. Adsorption is<br />
therefore achieved with a very short contact time (80% of the adsorption is achieved in<br />
< 3 min.) between the silica <strong>and</strong> the beer. This also provides the opportunity to treat only<br />
a proportion of the beer in a maturation vessel, i.e., that proportion destined for a long<br />
shelf-life (> 40 weeks). In this situation the filtration quality of the gel is as important as<br />
its adsorbent property. The best filterability is obtained with large size gel particles but<br />
these are not so good for protein adsorption. Generally, hydrogels are good adsorbents but<br />
because of small particle diameter (< 10 m) are relatively poor filter aids. Xerogels have<br />
larger diameter gel particles (up to 100 m) <strong>and</strong> so are less good at adsorption than<br />
hydrogels but have superior filtration characteristics. Rates of addition are normally<br />
between 50 <strong>and</strong> 80 g/hl. Brewers have to make a choice of which material to use<br />
depending on their circumstances. It is claimed (McKeown <strong>and</strong> Earl, 2000) that it is now<br />
possible to manufacture small particle sized xerogels that will deliver high filtration rates<br />
as well as displaying the optimum adsorption characteristics of a small sized hydrogel<br />
(< 10 m).<br />
Beer proteins are associated with the good foaming potential of beer. There has<br />
therefore been concern over the effects of silica gels on foam <strong>and</strong> head retention. The<br />
amino acid sequence of the protein determines its hydrophilic/hydrophobic balance. It is<br />
generally accepted that hydrophilic proteins are haze-formers <strong>and</strong> hydrophobic proteins<br />
promote foam.<br />
The selectivity of different silica gel preparations for removing haze-forming proteins<br />
has been studied in both all-malt <strong>and</strong> high adjunct (maize) beers (Guzman et al., 1999).<br />
Total protein adsorption was influenced by beer type but adsorption of haze-forming<br />
protein was not. Further, it was possible to manufacture a xerogel which showed<br />
selectivity for haze-forming, hydrophilic proteins. Maximum uptake appeared to occur<br />
around 100 m pore diameter.<br />
An essentially different method of using the adsorptive properties of silica derivatives<br />
to remove protein is to use the compound in the form of the liquid silica hydrosol, the<br />
intermediate in the manufacture of the hydrogel or the xerogel (Green et al., 2000). The<br />
hydrosol is a bluish opalescent liquid with a specific area of 300 m 2 /g in relation to its<br />
SiO 2 content. Essentially, if added to beer the hydrosol forms a gel by cross-linking the<br />
SiO 2 particles. The gel flocculates <strong>and</strong> sediments <strong>and</strong> so adsorbs protein, which settles<br />
out as tank bottoms. The silica sol can be added to the beer in the maturation tank.<br />
Careful mixing is required <strong>and</strong> considerable tank bottoms are produced which must be<br />
drawn off before filtration or removal by centrifugation. The sol can also be added with<br />
the body feed to the filter when it is effective in lowering haze but will not contribute to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
filtration performance like the gel. Further development work on silica sol <strong>and</strong> gel<br />
structure in relation to the removal of hydrophilic haze-forming protein from beer seems<br />
likely to be worthwhile. This technique remains very useful <strong>and</strong> has wide acceptance by<br />
brewers.<br />
15.3.3 Removal of polyphenols<br />
There is still some concern from brewers that techniques relying solely on the removal of<br />
protein for haze stabilization will affect beer foam. Techniques to remove polyphenols<br />
from beer have also been developed. The polyphenols responsible for haze-formation by<br />
interacting with proteins derive from malt with asmall contribution from hops. Thus the<br />
possibility exists of eliminating polyphenols from beer by eliminating them from malt.<br />
Two basic methods of polyphenol removal have therefore been practised; use of<br />
polyphenol-free malt <strong>and</strong> the much more widespread adsorption technique applied to<br />
wort or beer.<br />
Adsorption<br />
Early work involved the use of synthetic materials such as `Nylon 66', which were<br />
effective polyphenoladsorbents.FurtherworksawthedevelopmentofthepolymerPVPP<br />
(polyvinylpolypyrrolidone). This is across-linked polymer, which is insoluble in water,<br />
alcohol <strong>and</strong> acid (Gopal <strong>and</strong> Rehmanji, 2000) <strong>and</strong> hence has ahigh surface area for<br />
adsorption of haze-forming polyphenols. This occurs at the surface of the material by<br />
strong hydrogen bonding. It is proposed that the structure of the synthetic polymer limits<br />
internal bonding thus maximizing the number of external sites <strong>and</strong> reducing the<br />
concentration of PVPP needed for effective stabilization.<br />
PVPP can be employed as a`single use agent' <strong>and</strong> in this situation the insoluble<br />
PVPP-polyphenol complex is removed on the filter with kieselguhr depth filtration.<br />
PVPP for single use is ground to afine powder with alarge surface area to weight ratio<br />
<strong>and</strong> is used at rates of 15±25g/hl. The PVPP can be added to the beer stream along with<br />
kieselguhr at 0ëC (32ëF) <strong>and</strong>, like silica gel, ashort contact time is effective, although in<br />
this case about ten minutes is needed from the point of contact with the beer to removal<br />
on the filter. Some brewers prefer to add single use PVPP to the maturation tank. This is<br />
usually at rates of 10±15g/hl <strong>and</strong> alarge dense grade of powder is used that promotes<br />
sedimentation.Athoroughmixingofthetankcontentsbypumpedrecirculationisneeded<br />
for optimum effectiveness. The bulk of the PVPP-polyphenol complex is removed from<br />
the base of the tank after sedimentation, with final removal of suspended matter on the<br />
filter.<br />
PVPP can also be used as aregenerable product. Washing with hot caustic soda<br />
solution breaks the PVPP-polyphenol bonds. Addition rate is normally 30±50g/hl. The<br />
powder is of larger particle size <strong>and</strong> has greater mechanical strength than that for single<br />
use. The usual technique is to use the regenerable PVPP in ahorizontal leaf filter or a<br />
c<strong>and</strong>le filter (Section 15.5). Apre-coat of PVPP of 1±2mm (0.04±0.08in.) depth is<br />
layered onto the screens in the filter <strong>and</strong> PVPP is dosed into the beer stream by a<br />
proportioning pump. A rate of 10 hl/m 2 /h can be achieved <strong>and</strong> the precise rate is matched<br />
to that of the kieselguhr filter which follows. The treatment run will end when PVPP has<br />
filled the space between the screens of the filter. Spent PVPP is regenerated by<br />
circulating a solution of 1±2% sodium hydroxide at 60±80 ëC (140±175 ëF) through the<br />
PVPP filter bed for 15±30 minutes. The filter cake is then flushed with water at 80 ëC<br />
(175 ëF) to lower the pH value.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
This is normally followed by a rinse with dilute acid, usually phosphoric (1% w/w) or<br />
nitric (0.3% w/w), until the pH value of the solution leaving the filter is about 4.0. This<br />
treatment removes carbohydrates <strong>and</strong> calcium salts trapped in the debris of the filter aid.<br />
The filter is then washed with cold water to achieve a neutral effluent. Finally water <strong>and</strong><br />
carbon dioxide gas are used to displace the regenerated PVPP to the dosing tank. This can<br />
be assisted by spinning the screens in the filter to provide a centripetal force. Process<br />
losses are about 1% <strong>and</strong> these are made up by adding more PVPP to the dosing tank<br />
before repeating the cycle of treatment. There is greater capital cost in the use of<br />
redeemable PVPP but, for breweries with large outputs of beer requiring stabilization for<br />
long shelf-life (> 40 weeks), this is a cheaper option than single-use PVPP.<br />
Generally, it has been accepted that the most effective stabilization occurs, whether<br />
with PVPP or silica gel when combined with the use of temperatures of 0 to 2 ëC (32±<br />
28 ëF) <strong>and</strong> at times in excess of 48 hours. However, recent work suggests that PVPP can<br />
deliver stabilization at 4 ëC (39 ëF) which is equivalent to that obtained at 0 ëC when<br />
assessed by shelf-life measurement (Byrne et al., 1999). This could be of considerable<br />
significance in energy saving.<br />
Proanthocyanidin free malt<br />
Polyphenols in beer in the main derive from malt. The most reactive of the polyphenols<br />
are proanthocyanidins or anthocyanogens <strong>and</strong> are found in the barley testa. It would<br />
therefore seem feasible that if malt could be produced from a barley free of<br />
anthocyanogens then haze stability might be achieved without the intensive postfermentation<br />
treatments described above.<br />
Barley breeding work was carried out in the Carlsberg laboratories in the early 1970s<br />
with the objective of inducing mutations in barley so that anthocyanogen producing genes<br />
were ineffective (Jende-Strid, 1997). It should be noted that the resulting barleys were not<br />
produced by genetic transformation by the transplanting of genes from one species to<br />
another. The induced mutations were transmitted to commercial cultivars by classical<br />
techniques of crossing <strong>and</strong> re-selection. In the last 30 years over 700 mutants of winter<br />
<strong>and</strong> spring barleys have been produced (Sole, 2000). The problem has been the<br />
commercial acceptability of the varieties from both an agronomic <strong>and</strong> malting <strong>and</strong><br />
brewing st<strong>and</strong>point. Some varieties showed no dormancy <strong>and</strong> so lodged <strong>and</strong> effectively<br />
malted in the field <strong>and</strong> so were useless for commercial malting. There was no doubt<br />
however that beers could be brewed from anthocyanogen-free malts, which had haze<br />
shelf-lives equivalent to those beers stabilized with PVPP <strong>and</strong>/or silica gel.<br />
The latest varieties to be trialled are Caminant, Chamant <strong>and</strong> Prominant in mainl<strong>and</strong><br />
Europe <strong>and</strong> Clarity in the UK. There has been relatively little interest in other parts of the<br />
world. Clarity appears to have sound agronomic properties <strong>and</strong> has been grown<br />
successfully on the light l<strong>and</strong>s of eastern Engl<strong>and</strong>. Clarity malt has been trialled in pilot<br />
breweries at <strong>Brewing</strong> Research International in the UK <strong>and</strong> at commercial breweries.<br />
Advantages have been demonstrated over conventional malts in lower polyphenol levels<br />
in wort <strong>and</strong> in haze shelf-life tests (Sole, 2000) <strong>and</strong> in the opportunity to use higher<br />
temperatures for stabilization (+4 ëC compared to 1 ëC). The problem has been in lower<br />
extract yield in the brewhouse, which has often been at least 1%, <strong>and</strong> this has limited<br />
acceptability of Clarity malt for some brewers. This has been counteracted in some<br />
situations by using Clarity malt as a proportion of the grist (say 50%).<br />
Full acceptance of the use of anthocyanogen-free malts as an effective means of<br />
stabilizing beer awaits the demonstration of the economics of the whole system in<br />
relation to the costs of PVPP treatment.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
15.3.4 Combined treatments to remove protein <strong>and</strong> polyphenols<br />
Refined silica gel treatments are very effective at selectively removing haze-forming<br />
protein <strong>and</strong> PVPP will effectively remove the haze-forming polyphenols from beer.<br />
However,there is stillsome riskthat foam potential can be reduced byexcessive removal<br />
of protein <strong>and</strong> there is abelief that polyphenols contribute anti-oxidant properties to beer<br />
<strong>and</strong> give improved texture to beer in the mouth. These considerations have been used as<br />
reasons to favour astabilization approach dependent on the removal of defined amounts<br />
of both protein <strong>and</strong> polyphenol. PVPP can be used with papain (Esnault, 1995) <strong>and</strong> is<br />
added with the enzyme. The dangers of excessive proteolysis with papain remain. PVPP<br />
cannot be added with gallotannins as it complexes immediately <strong>and</strong> will not adsorb the<br />
polyphenols in the beer. It can be added after gallotannin treatment (Musche <strong>and</strong> de<br />
Pauwe, 1999) when it will adsorb polyphenols <strong>and</strong> will complex residual tannic acid that<br />
can contribute aharsh astringency to the beer.<br />
The most effective combined treatments have seen PVPP used along with silica gel<br />
added from separate dosing vessels (McMurrough et al., 1999). Combined stabilization<br />
was shown to be more cost effective than the intensive treatments needed when PVPP or<br />
silica gel are used alone. Recently acombined PVPP/silica xerogel (3:7) preparation has<br />
been described (Gopal et al., 1999) that works to reduce proteins <strong>and</strong> polyphenols in a<br />
single addition at 80g/hl. This treatment is used without PVPP regeneration <strong>and</strong> achieves<br />
stabilization at lower dose rates than the treatments used alone which required silica gel<br />
at 95g/hl <strong>and</strong> PVPP at 20g/hl. Filtration characteristics on the kieselguhr filter were also<br />
excellent <strong>and</strong> better than hydrogels used alone.<br />
The relative costs of treatments to remove protein <strong>and</strong> polyphenols from beer vary<br />
over time. Systems dependent on single use PVPP have high revenue costs. But if small<br />
volumes are being treated then the capital cost of recovery systems is avoided. Silica<br />
xerogels were expensive but recent developments have seen prices falling. The most<br />
revenue cost effective treatment is probably gallotannin treatment combined with<br />
recoverable PVPP, but this will almost certainly require the use of acentrifuge to process<br />
tank bottoms <strong>and</strong> there is considerable capital cost in the recovery system. In the final<br />
assessment of which treatment to use, cost will be only one factor. Brewers will rely on<br />
their experience of the ways their beers behave to achieve the maximum stability<br />
required.<br />
15.3.5 Hazes from other than protein or polyphenols<br />
Haze in beer can sometimes be caused by carbohydrate or bacterial cell wall material or<br />
calcium oxalate. Attempts to remove these hazes at the maturation stage are usually<br />
failures. These hazes do not respond to stabilization treatments appropriate for proteins/<br />
polyphenols. Beer hazes derived from carbohydrates or oxalate certainly can be<br />
controlled or eliminated by attention to raw materials or the conditions of mashing.<br />
Sufficient calcium must be present in the wort to precipitate oxalate from the malt or<br />
other grist materials.<br />
Theachievementofoptimumstarchconversiondependsonadequateamylaselevelsin<br />
the maltbut moreespecially onthe attainment ofgelatinization temperatures inthe mash.<br />
This should result in no problems with -glucan hazes in beer. Alow-temperature st<strong>and</strong><br />
in the mash (
contamination during germination. This should be avoided with proper specification <strong>and</strong><br />
dealing with reputable maltsters. All these aspects are discussed in other parts of this<br />
book.<br />
15.4 Carbonation<br />
Carbon dioxide is avery important constituent of beer. It imparts sparkle <strong>and</strong> `mouth<br />
feel' <strong>and</strong> sharpness associated with its properties as an acid gas. The concentration of<br />
carbondioxide inbeerforsaleiscarefully controlledtoensurethatconsumers ofthebeer<br />
can drink aconsistent product. Beers that lack carbon dioxide, particularly lager beers,<br />
are dull <strong>and</strong> lifeless <strong>and</strong> are said to lack condition <strong>and</strong> be flat. The carbon dioxide is the<br />
gas produced naturally in primary <strong>and</strong> secondary fermentation <strong>and</strong> that added to the beer<br />
by `carbonation'.<br />
At the end of primary fermentation the concentration of carbon dioxide in beer can<br />
vary from about 2g/l (1 vol/vol) in ashallow ale fermenting vessel up to 5g/l (2.5 vol/<br />
vol) in adeep cylindroconical vessel. During secondary fermentation the level of carbon<br />
dioxide will increase. If acontrolled secondary fermentation is not carried out then<br />
carbon dioxide will normally have to be added to the beer. Such is the control now<br />
dem<strong>and</strong>ed for the precise level of carbon dioxide in finished beer that the gas levels are<br />
frequently adjusted after primary <strong>and</strong>/or secondary fermentation. The amount of carbon<br />
dioxide that will dissolve in beer depends on temperature <strong>and</strong> pressure.<br />
15.4.1 Carbon dioxide saturation<br />
Enclosed fermenters <strong>and</strong> maturation vessels are fitted with automatic gas pressure<br />
regulators that can be set at apredetermined pressure. Assuming that the gas above the<br />
beer is pure carbon dioxide then the gas that will dissolve in the beer at equilibrium is<br />
shown in Fig. 15.4. It can be seen that at agiven temperature <strong>and</strong> pressure aparticular<br />
equilibrium condition willbe reached. Increasing pressure willlead toalinear increasein<br />
the weight of carbon dioxide dissolving in the beer. An increase in temperature will give<br />
anon-linear decrease in the amount of gas dissolved (Fig. 15.5). These factors derive<br />
from Henry's Law, which states that the concentration of gas in the liquid phase is equal<br />
to the imposed pressure on the gas divided by Henry's constant that is temperature<br />
dependent. This is of fundamental importance in practical brewing <strong>and</strong> it must be<br />
remembered that as temperature rises substantially less carbon dioxide will dissolve in<br />
beer. It is therefore extremely important to keep beer as cold as is practicable after<br />
fermentation to keep carbon dioxide in solution <strong>and</strong> avoid the necessity for excessive<br />
artificial carbonation. It is normally the convention in brewing to state the concentration<br />
of carbon dioxide as volumes of gas at st<strong>and</strong>ard temperature <strong>and</strong> pressure per volume of<br />
beer. A gram molecule of a perfect gas occupies 22.4 litres at STP <strong>and</strong> so one volume of<br />
carbon dioxide is equivalent to 0.196% carbon dioxide by weight.<br />
At constant temperature <strong>and</strong> pressure the amount of carbon dioxide dissolving in beer<br />
will be a function of the time of contact between beer <strong>and</strong> gas <strong>and</strong> will decrease<br />
exponentially as equilibrium is approached. Increasing the surface of the beer exposed to<br />
gas can increase the rate of dissolution <strong>and</strong> a very shallow layer of beer will approach<br />
saturation more quickly. Further, if bubbles are created then a very large surface area is<br />
presented for gas transfer into the surrounding beer. In deep tanks, hydrostatic pressure<br />
also affects the concentration of carbon dioxide in beer (about 0.5 vol. increase/metre<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Pressure (psig)<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
Pressure (bar)<br />
4<br />
3<br />
2<br />
1<br />
0<br />
0 5 10 15 20 25 30<br />
Temperature (°C)<br />
35 40 45 50 55 60 65 70 75 80 85<br />
Temperature (°F)<br />
Fig. 15.4 Relationship between equilibrium values for dissolved carbon dioxide, temperature <strong>and</strong><br />
pressure (Hough et al., 1982).<br />
Pressure (psig)<br />
80<br />
60<br />
40<br />
20<br />
10<br />
Pressure (bar)<br />
5<br />
4<br />
3<br />
2<br />
1<br />
depth) <strong>and</strong> the concentration at the base of the tank is higher than at the surface.<br />
Convection currents in the beer affect this gradation.<br />
Beer is capable of holding carbon dioxide in a supersaturated state, so rapid release of<br />
pressure or increase in temperature does not immediately lead to attainment of the<br />
appropriate equilibrium. This is advantageous when serving highly carbonated beers (say<br />
3 vol. CO 2) from bottles. When this beer is poured into a glass, the gas does not normally<br />
effervesce uncontrollably (`gushes') but releases carbon dioxide slowly. Knowledge of<br />
3.0<br />
2.5<br />
2.0<br />
1.8<br />
1.4<br />
1.0<br />
4 vol. 3 vol.<br />
Carbon dioxide content (volumes)<br />
10 20 30 40 50<br />
Temperature (°C)<br />
40 50 60 70 80 90 100 110 120<br />
Temperature (°F)<br />
2 vol.<br />
1 vol.<br />
Fig. 15.5 Effect of temperature on pressure for fixed contents of carbon dioxide. Note that an<br />
increase in temperature gives a non-linear decrease in the amount of gas that dissolves, viz., say at 3<br />
bar the equilibrium occurs for 4 vol. at 15 ëC, 3 vol. at 26 ëC <strong>and</strong> for 2 vol. at 41 ëC (Hough et al.,<br />
1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
0.6<br />
0.4<br />
0.2<br />
Carbon dioxide content (% w/v)
Henry's Law thus provides the underst<strong>and</strong>ing for controlling the carbon dioxide content<br />
of the beer prior to filtration <strong>and</strong> subsequent packaging. For st<strong>and</strong>ard lager beers it is<br />
normal to aim for around 5g/l (2.5 vol. or 0.5%) of carbon dioxide in the beer. The<br />
pressure release valve will normally be set to give aslightly higher carbon dioxide<br />
content to compensate for the subsequent losses in filtration <strong>and</strong> packaging.<br />
Inevitably, some carbon dioxide produced during secondary fermentation will escape.<br />
This escaping gas is an important factor in beer maturation because it exerts acleansing<br />
effect by purging volatile fermentation products <strong>and</strong> so accelerating the flavour<br />
improvement process.<br />
15.4.2 Carbon dioxide addition<br />
All processes after secondary fermentation should be designed to keep carbon dioxide in<br />
solution in the beer. Thus beer should be kept cold <strong>and</strong> under the appropriate pressure of<br />
carbon dioxide to prevent gas release. However, situations will arise when temperature<br />
will rise or pressure will fall <strong>and</strong> gas will escape. The carbon dioxide must then be<br />
replaced in the beer before packaging. This is frequently achieved after filtration during<br />
transit to the bright beer tanks.<br />
This addition of carbon dioxide can occur whilst chilling beer through aplate heat<br />
exchanger (Fig. 15.6) <strong>and</strong> so can take advantage of the turbulence of the beer in creating<br />
good conditions for gaseous exchange. Apurpose-designed carbonation unit (Fig. 15.7)<br />
can also be used. This consists of a long pipe usually in the form of U-tube bends through<br />
which the beer flows. Carbon dioxide is injected as fine bubbles <strong>and</strong> the uptake, even in<br />
this form, can take a considerable time. The carbon dioxide must be the purest form<br />
available <strong>and</strong> no oxygen must be introduced. The injection unit must be easy to clean <strong>and</strong><br />
must be cleaned regularly. Carbon dioxide can also be added `in-vessel' but this is<br />
frequently less efficient <strong>and</strong> more difficult to control. A `carbonation stone' is sometimes<br />
used to ensure production of fine bubbles of carbon dioxide to aid dissolution in the beer.<br />
This technique is sometimes described as `gas washing' <strong>and</strong> provides an opportunity for<br />
the removal of oxygen <strong>and</strong> unwanted flavour volatiles as well as carbonation. After<br />
`washing' the vessel must be sealed to allow pressure build-up <strong>and</strong> the dissolution of<br />
carbon dioxide. There are a number of problems associated with externally added carbon<br />
dioxide <strong>and</strong> it is good <strong>practice</strong> to avoid this technique as far as possible.<br />
Care must also be taken not to dissolve more carbon dioxide in the beer than the<br />
specification allows. Reducing carbon dioxide levels is difficult without creating froth or<br />
fob. This reduction can be achieved by gas washing with careful bubbling of oxygen-free<br />
Conditioning<br />
tank<br />
Coolant<br />
CO 2<br />
Plate heat<br />
exchanger<br />
chiller<br />
Cold tank<br />
Fig. 15.6 Carbonation <strong>and</strong> chilling of beer `on the run', from conditioning tank to cold tank<br />
(Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
a<br />
Beer<br />
b<br />
e<br />
c<br />
nitrogen gas through the beer. Nevertheless, this technique is often a source of<br />
reintroduction of oxygen to beer with adverse effects on haze <strong>and</strong> flavour stability <strong>and</strong><br />
loss of foaming potential.<br />
It mustnever be assumed that over carbonation of abeer reduces the uptake of oxygen<br />
if the beer is exposed to air. This uptake will be afunction of the difference in partial<br />
pressures between the beer <strong>and</strong> the air in contact with it <strong>and</strong> is not influenced by the<br />
carbon dioxide content of the beer. Turbulence in movement should be avoided for beer<br />
which has been carbonated, either naturally or by external addition. Turbulence will lead<br />
to the loss or `break-out' of carbon dioxide, the potential pick-up of oxygen <strong>and</strong> loss of<br />
foam potential because of protein denaturation at the gas/liquid interfaces.<br />
15.4.3 Carbon dioxide recovery<br />
The cost of carbon dioxide gas varies but it is generally expensive <strong>and</strong> certainly costs<br />
more than nitrogen. Before contemplating the economics of carbon dioxide recovery in a<br />
brewery the use of nitrogen gas as an alternative tocarbon dioxide for ensuring anaerobic<br />
h<strong>and</strong>ling of beer should be maximized. On this basis the amount of carbon dioxide<br />
required in breweries varies considerably from about 1.3 to 2.0kg of carbon dioxide/hl of<br />
beer produced. Of course carbon dioxide is essential for many purposes <strong>and</strong> there is the<br />
opportunity to collect the gas from enclosed vessels during fermentation. An essential<br />
aspect of recovery systems is that the gas must be purified if it is to be added to beer.<br />
Carbon dioxide is normally collected through afob tank into an inflatable balloon<br />
(Fig. 15.8), <strong>and</strong> from there it passes through water scrubbers <strong>and</strong> carbon purifiers before<br />
being compressed at the liquefying pressure of 18±22 bar (265 to 320lb./in. 2 ).This<br />
creates considerable heat <strong>and</strong> the compressed gas must then be cooled <strong>and</strong> so liquefied.<br />
The gas is dried through alumina driers <strong>and</strong> then stored as liquid until required for use<br />
when it can be restored to the gaseous state through an evaporator. Up to 2.0kg of carbon<br />
dioxide per hl of beer can be recovered by this technique <strong>and</strong> this could represent the<br />
whole of abrewery requirement. However, recovery is often much less than this as a<br />
resultoflosses<strong>and</strong>cleaningofthegas.Manybrewersexpectrecoverysystemstoprovide<br />
around 60% of requirement.<br />
d<br />
Beer<br />
f CO 2<br />
Fig. 15.7 Carbonation unit; (a) venturi nozzle, (b)carbon dioxide control valve,(c) CIP valve, (d)<br />
carbon dioxide sensor, (e) carbon dioxide dissolving area, (f) carbon dioxide measuring device<br />
(Kunze, 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
FV<br />
Fob tank<br />
CO 2 compressor<br />
control switch<br />
Steam<br />
CO 2<br />
evaporator<br />
Condensate<br />
CO 2 gas<br />
Balloon<br />
Liquid CO 2<br />
tank<br />
Excess<br />
pressure<br />
device 0–3 bar<br />
Ammonia<br />
compressor<br />
<strong>and</strong> cooler<br />
Booster<br />
compressor<br />
15°C<br />
Pump<br />
Water cooler<br />
Twin activated<br />
alumina driers<br />
Water<br />
scrubber<br />
17 bar<br />
Water<br />
Compressor<br />
<strong>and</strong><br />
intercooler<br />
Fig. 15.8 Equipment for collecting, purifying, storing, <strong>and</strong> releasing carbon dioxide (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Twin activated<br />
carbon purifiers
15.5 Clarification <strong>and</strong> filtration<br />
Clarification of beer involves the removal of yeast <strong>and</strong> the sedimented protein <strong>and</strong><br />
polyphenol haze material derived from the beer stabilization techniques <strong>and</strong> cold break<br />
(Section15.3). Beerflavour is considerably morestablewhenit containssuspended yeast<br />
as the yeast promotes strongly reducing conditions. Total removal of yeast should<br />
therefore be delayed to the last possible moment before packaging. In cask beer yeast is<br />
never totally removed <strong>and</strong> so this beer is particularly robust with respect to flavour<br />
stability. There is, however, something of adilemma here because to achieve satisfactory<br />
filtration, beer going to the filter must contain
days from the end of primary fermentation. As noted, considerable effort has been<br />
directed to accelerating the flavour maturation of beer <strong>and</strong> the development of haze<br />
stability. The separation of yeast cannot be allowed to become rate limiting.<br />
Considering Stokes' Law from the viewpoint of sedimentation of yeast in beer, it<br />
would be difficult to lower the viscosity of the solution without a detrimental effect on<br />
beer quality. Likewise to create a marked difference between the density of the particles<br />
<strong>and</strong> the surrounding beer could alter the path of flavour maturation. An obvious method<br />
to improve the rate of sedimentation would be to increase the value of the factor 2r 2 , i.e.,<br />
to increase the size of the yeast cells by agglomeration, i.e., causing them to clump<br />
together. This can be achieved by the use of isinglass finings. Isinglass finings are<br />
prepared from the dried swim bladders of certain fish (often sturgeon), which live in the<br />
estuaries of tropical rivers such as the Mekong <strong>and</strong> the Amazon (Leach <strong>and</strong> Barrett,<br />
1967). To cope with large changes in water density as a result of experiencing fresh <strong>and</strong><br />
salt water conditions alternately these fish develop very large bladders to adjust their<br />
buoyancy. These bladders are composed of almost pure collagen <strong>and</strong> they are one of the<br />
purest forms of protein found in nature. An isinglass solution is obtained by soaking the<br />
dried bladders in dilute solutions of cold tartaric, sulphurous <strong>and</strong> sometimes malic acid<br />
for periods up to six weeks (a process known as cutting). A turbid, colourless, viscous<br />
solution results, which contains soluble collagen, gelatine, (the denaturation product of<br />
collagen) <strong>and</strong> some insoluble material. Only the soluble collagen contributes to fining<br />
action <strong>and</strong> methods are available for determining the collagen content of fining solutions<br />
(Leach <strong>and</strong> Barrett, 1967).<br />
Effective fining is achieved as a result of the positive charge on the collagen molecule<br />
that exists as a triple helix of complex stereochemistry rich in basic amino acids. Yeast<br />
cells are thought to flocculate as a result of surface proteins (lectins) bridging with<br />
carbohydrate side chains (mannan) on neighbouring cell walls (Shiel, 1999). It has been<br />
suggested that isinglass functions as an extracellular lectin with the carbohydrate moieties<br />
on yeast cell walls bearing a negative charge bridging to -NH3 + sites on the collagen<br />
molecule. In this way yeast cells will clump together <strong>and</strong> so the factor 2r 2 in Stokes' Law<br />
will be enhanced with a corresponding increase in sedimentation rate. Breweries can use<br />
isinglass either as a liquid purchased from the manufacturer or as shredded swim bladders<br />
ready for cutting, with acid, in the brewery.<br />
Larger molecules of soluble collagen are likely to be more effective at fining than<br />
smaller molecules. Large molecules will have more charged sites per molecule <strong>and</strong><br />
therefore will react more readily with <strong>and</strong> cross-link the negatively charged yeast cells.<br />
Finings from different sources do not differ much with respect to total charge or charge<br />
distribution but the constituent collagen molecules do differ in size <strong>and</strong> size distribution.<br />
The size distribution is a function of the ease with which collagen is broken down by acid<br />
<strong>and</strong> the extent to which aggregates of molecules are held together by weak physical <strong>and</strong><br />
chemical forces.<br />
The collagen triple helix forms a long rigid rod of 1.5 nm diameter <strong>and</strong> the individual<br />
chains exhibit frequent sharp twists owing to the presence of adjacent imino acids,<br />
proline <strong>and</strong> hydroxyproline. As a result of the high hydroxyl group content there is strong<br />
hydrogen bonding between the three chains. Young tissues contain monomers of 300 nm<br />
length <strong>and</strong> 300,000 molecular weight.<br />
The molecular size of the soluble collagen can be estimated by measuring the intrinsic<br />
viscosity (Leach <strong>and</strong> Barrett, 1967):<br />
‰ Š ˆ KM<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 15.1 Comparison of fining properties of isinglass from different sources (proportion of<br />
effective finings derived from isinglass ˆ A B=100%)<br />
(Leach <strong>and</strong> Barrett, 1967)<br />
Type of leaf Form Soluble Soluble Intrinsic<br />
nitrogen (%) collagen (%) viscosity [ ]<br />
A B (dl/g)<br />
Karachi Flock 80 86 15.4<br />
Karachi Shredded 78 96 21.9<br />
Brazil Lump Flock 81 97 18.2<br />
Brazil Lump Shredded 84 93 18.7<br />
Long Saigon Flock 62 87 16.5<br />
Long Saigon Shredded 57 92 2.6<br />
Round Saigon Shredded 97 94 26.5<br />
Penang Shredded 20 98 16.8<br />
where Kis aconstant depending on the particular solute/solvent system, Mis the average<br />
molecular weight <strong>and</strong> is aconstant dependent on molecular shape <strong>and</strong> rigidity, for<br />
isinglass is approximately 2<br />
\ p ˆKM<br />
Intrinsic viscosities are independent of concentration <strong>and</strong> are high for finings<br />
compared to proteins in general. If beer is easy to fine then in citrate solution [ ]should<br />
be >16 <strong>and</strong> it is found that different types of isinglass show little difference in fining<br />
action. Beer that is more difficult to fine needs finings where [ ]is >20 <strong>and</strong> then<br />
differences are manifest between different forms of isinglass (Table 15.1). Collagen<br />
rapidly denatures to inactive gelatine when the temperature rises <strong>and</strong> fining solutions<br />
must be stored at temperatures of 4±10ëC (39±50ëF). Finings can be used effectively at<br />
temperatures up to 14ëC (57ëF) <strong>and</strong> work best when the temperature of the beer is rising<br />
slightly. This particularly applies to the use in cask beer when the finings are normally<br />
added at racking prior to despatch from the brewery (Chapter 21).<br />
The use of isinglass finings in the preparation of brewery conditioned beer is not<br />
universal. It is largely a UK <strong>practice</strong> deriving from its effectiveness for cask beer. There<br />
is, however, renewed interest in the USA <strong>and</strong> the use of finings is increasing in Australia<br />
<strong>and</strong> Africa. Isinglass finings improve foam stability of beer by removing lipid material<br />
such as fatty acids <strong>and</strong> phospholipids, which are foam negative. This is an important<br />
secondary characteristic. The use of isinglass along with silica hydrogel has also recently<br />
been described (Shiel, 1999) <strong>and</strong> improvements in filter runs (by as much as threefold)<br />
<strong>and</strong> actual beer shelf-life were demonstrated. This seems likely to be as a result of the<br />
huge removal by isinglass of particles < 4.5 m in diameter that will cause blockage of the<br />
filter <strong>and</strong> have the potential to cause haze in the beer.<br />
Sedimentation of yeast in the presence of isinglass is rapid <strong>and</strong> compact easily<br />
removable tank bottoms are formed. This may require further processing by<br />
centrifugation or filtration to recover entrapped beer <strong>and</strong> finally separate the yeast if<br />
this is deemed economically sensible. It is clearly established (Shiel, 1999) that isinglass<br />
is completely removed during the brewing process <strong>and</strong> is not detectable in finished beer.<br />
Centrifugation<br />
The other powerful component of Stokes' Law is the effect of the acceleration owing to<br />
gravity. Centrifuges increase the gravitational force <strong>and</strong> so will allow more rapid<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
separation <strong>and</strong> removal of the yeast <strong>and</strong> other particles. Centrifuges are often used where<br />
isinglass finings are not used or they are used in conjunction with isinglass to separate<br />
yeast <strong>and</strong> beer in tank bottoms. A centrifuge operating at 5,000 rpm can effect an 8,000<br />
increase in the gravitational force. The general relationship being:<br />
g ˆ ! 2 r<br />
where g is the gravitational force, ! is the rotational speed in radians/second <strong>and</strong> r is the<br />
radius in metres.<br />
Sedimentation of the particles is also enhanced if the settling distance is reduced <strong>and</strong><br />
this is an important feature of centrifuges. Centrifuges were developed from<br />
considerations of the behaviour of solids <strong>and</strong> liquids, <strong>and</strong> liquids of different densities<br />
in balanced tanks. If such a balanced tank is rotated about its vertical axis centrifugal<br />
force supplements the acceleration owing to gravity. Discs can be inserted in the<br />
centrifuge to reduce the path distance of sedimentation. Two main types of centrifuge are<br />
now used in breweries for the separation of yeast from beer prior to filtration, selfcleaning<br />
clarifiers <strong>and</strong> decanter clarifiers. Self-cleaning clarifiers can operate at relatively<br />
low solids contents to process complete tanks of beer, say between fermentation <strong>and</strong><br />
maturation, when other means of yeast separation such as isinglass finings <strong>and</strong><br />
sedimentation are not available or are not fast enough. They do not operate well on tank<br />
bottoms where the solids content can be very high. In this situation the decanter machine<br />
would be used.<br />
· Self-cleaning clarifiers. In this type of centrifuge the solids are discharged at intervals<br />
whilst the centrifuge is operating at full speed (Fig. 15.9). The bowl contains discs<br />
separated by spaces of 0.5±2.0 mm (0.02 to 0.08 in.) by distance pieces known as<br />
`chaulks'. The yeast <strong>and</strong> other solids slide on the disc surface to the periphery <strong>and</strong> collect<br />
on the surface of the bowl. Clarified beer moves towards the centre of the machine <strong>and</strong> is<br />
pumped to the next processing stage, which is normally the maturation vessel when the<br />
centrifuge is inserted in the line between fermentation <strong>and</strong> maturation vessels.<br />
Depending on the solids load the rate of operation can vary from 40 to 600 hl/h. Yeast<br />
is removed by a mechanism whereby the spinning bowl separates momentarily into two<br />
parts at the rim <strong>and</strong> the solids are ejected. This ejection can be on a simple predetermined<br />
time basis, irrespective of the presence of sludge, in which case beer losses<br />
can be high. The self-sensing type of machine uses a hydraulic differential pressure<br />
Solids<br />
Feed<br />
Clarified liquid<br />
Solids<br />
Fig. 15.9 Self-cleaning clarifier centrifuge (Hough, et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Rotor<br />
drive<br />
Outlet for<br />
liquid<br />
Scroll<br />
Outlet for<br />
solids<br />
mechanism for sensing the accumulation of solids in the bowl. These machines offer<br />
better control of beer losses but must be constantly maintained. A third system works by<br />
optically monitoring the clarity of the beer leaving the machine.<br />
· Decanter clarifiers. If the solid content of the beer to be processed is very high, as<br />
can be the case in tank bottoms (up to 60% by volume), a decanter centrifuge can be<br />
used (Fig. 15.10). These machines employ a rotating screw in a casing with discharge<br />
of solids at one end <strong>and</strong> processed beer at the other. They normally operate at speeds<br />
of 40 hl/h.<br />
Centrifuges have a number of advantages for yeast removal <strong>and</strong> clarification of beer.<br />
They have a small space requirement <strong>and</strong> can be sterilized <strong>and</strong> maintained sterile. They<br />
can be hermetically sealed <strong>and</strong> this is essential to exclude oxygen. Fitting seals achieves<br />
this where the rotating parts of the machine adjoin the stationary parts. Centrifuges have<br />
no requirement for filter aids <strong>and</strong> no active adsorption is involved. If a constant solids<br />
load is presented it is possible to operate a centrifuge continuously for an indefinite<br />
period. Yeast separated by centrifugation varies between 13 <strong>and</strong> 25% dry matter.<br />
Centrifuges have a very high requirement for electrical energy. There is a high inertia<br />
of the rotating parts <strong>and</strong> when slurry is discharged an equivalent volume of beer entering<br />
the machine is brought to rotational speed in a very short time. Motors have to be sized to<br />
meet these maximum loads, which on a 200 hl/h centrifuge would be 22.5 kW. The<br />
normal running rate of a centrifuge will consume about 0.35 MJ of energy per hl of beer<br />
processed. Centrifuges are noisy, frequently exceeding 85 dB(A) <strong>and</strong> so ear protection<br />
must be worn when inspecting or working on them. This sometimes deters maintenance<br />
<strong>and</strong> inspection. The machines are complex <strong>and</strong> difficult to maintain <strong>and</strong> spares are costly.<br />
These are real disadvantages in a modern brewery operating at low fixed cost. A further<br />
problem is the rise in temperature of the beer <strong>and</strong> yeast, which occurs during<br />
centrifugation (up to 3 ëC, 5 ëF). This can lead to a loss in carbon dioxide <strong>and</strong> a need to rechill<br />
the beer. The physiological condition of the yeast is adversely affected by shear<br />
forces <strong>and</strong> the temperature rise <strong>and</strong> yeast collected by centrifugation is far less suitable<br />
for re-pitching than that collected by natural sedimentation.<br />
Filtration<br />
Yeast can be separated from beer <strong>and</strong> beer recovered by various types of filtration: the<br />
yeast press, rotary vacuum filtration <strong>and</strong> cross-flow filtration. Filtration techniques are<br />
normally used for the processing of tank bottoms. Tank bottom beer (sometimes called<br />
`barm ale') may represent 1±3% of the total volume of the beer in the tank. It is thus a<br />
high process loss to discard the whole of this beer. This is often not of such good quality<br />
Feed<br />
Fig 15.10 Decanter centrifuge (based on drawing received from Alfa-Laval Ltd.).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
as the bulk of the beer in the tank. It usually has ahigher pH value <strong>and</strong> contains more<br />
amino nitrogen <strong>and</strong> higher alcohols. Recovered beer must be blended at some stage with<br />
primary tank beer. This should be carried out as early as is possible in the process as the<br />
recovered beer will contain yeast autolysate.<br />
The point of addition will vary from brewery to brewery depending on the available<br />
equipment <strong>and</strong> the quantity of beer. A good point of addition is the beginning of<br />
fermentation when the actively fermenting pitching yeast will quickly absorb the<br />
metabolic products of the autolysed yeast in the recovered beer with no deleterious effect<br />
onoverall beer quality. Earlier additions intothe brewhousearealsopossible.The capital<br />
cost of making these additions <strong>and</strong> indeed recovering the yeast may be high <strong>and</strong> the<br />
overall economics of the process must be assessed when deciding on the optimum<br />
method inaparticularbrewery.Inmanypartsofthedeveloped worldthecostsoftreating<br />
effluent containing yeast <strong>and</strong> beer are so high that it will be essential from afinancial<br />
viewpoint to separate yeast <strong>and</strong> recover beer from tank bottoms before discharging<br />
effluent to drain (Chapter 3).<br />
· Yeast press. In traditional ale brewing in the UK it was common <strong>practice</strong> to press<br />
yeast after skimming from the fermenter to separate it from barm beer. The pressed<br />
yeast was stored as adry cake in arefrigerator or was slurried in chilled water until<br />
reused. This was agood way to recover barm beer for addition back to the process <strong>and</strong><br />
the pressed waste yeast in aconvenient form to sell for the manufacture of yeast<br />
extracts or animal feed. Yeast was also sold to the distilling industry in this form.<br />
Yeast presses are made of gun metal <strong>and</strong> contain fabric cloths. Two cloths form a<br />
chamber on either side of a backing plate. The yeast slurry is pumped into the<br />
chambers <strong>and</strong> the beer recovered from channels in the backing plate. Presses are<br />
labour intensive <strong>and</strong> slow, requiring manual cleaning of cloths <strong>and</strong> two-man operation<br />
for yeast removal. In many breweries the traditional press has been discarded <strong>and</strong> the<br />
yeast is used <strong>and</strong> stored as barm (i.e. as aslurry in beer).<br />
There have been recent developments in yeast press design dem<strong>and</strong>ing a<br />
reconsideration of this technique for tank bottom processing. Inflatable diaphragm<br />
plates made of polypropylene have been introduced which can squeeze the cake <strong>and</strong><br />
considerably shorten the time of pressing. New presses can be in-place cleaned at<br />
80ëC (175ëF) <strong>and</strong> can be operated by one man. Afully automated press has also been<br />
described (Anderson et al., 2000) in which astack of vertical plates is used with a<br />
continuous polypropylene cloth driven by power rollers fed around each of the plates.<br />
In the modern press the pressure applied rises from 4to 19 bar (60±280lb./in. 2 )as<br />
the chamber is filled <strong>and</strong> kieselguhr or perlite is sometimes added to improve<br />
filterability. Therecoveredbeer cancontain0.1to0.5 10 6 yeast cells/ml ofbeer <strong>and</strong><br />
the concentration of the pressed yeast is between 25 <strong>and</strong> 40% dry weight. The market<br />
value of pressed yeast varies but it can normally be sold at aprofit. The beer is bitter<br />
<strong>and</strong> yeasty <strong>and</strong>contains high levelsofamino acids <strong>and</strong>polyphenols. It can be added at<br />
the start of fermentation but in some breweries it is pasteurized before addition. If<br />
possible the beer should be added after wort boiling but before wort cooling <strong>and</strong><br />
pitching. Experience shows that beer from ayeast press should not be added at a<br />
greater rate than 5%.<br />
· Rotary vacuum filter (Fig. 15.11). This machine can operate continuously <strong>and</strong><br />
consists of a rotating drum covered by a filter sheet on which is deposited a pre-coat of<br />
perlite or kieselguhr. Beer is sucked into the drum <strong>and</strong> yeast collects as a layer on the<br />
surface, which is then removed by knives to a tank underneath the filter. The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Outlet<br />
Part of tank<br />
receiving slurry<br />
Pipes collecting filtrate<br />
Precoat<br />
Septum<br />
Septum support<br />
concentration of the recovered yeast is usually 10 to 25% dry matter. The beer has<br />
similar properties, although it contains more oxygen, than that from the yeast press<br />
<strong>and</strong> can be returned to the process in asimilar way.<br />
· Cross-flow filter. In this system the yeast/beer mixture is circulated under pressure<br />
over amembrane. For the system to be successful shear at the membrane surface is<br />
necessary to prevent fouling <strong>and</strong> amechanism for the removal of solids is essential.<br />
Shear can be created by high volume flow at the membrane surface, but these systems<br />
have not always been commercially successful because of the high energy input<br />
needed <strong>and</strong> the high pumping rate leading to yeast cell damage <strong>and</strong> consequent<br />
adverse flavour effects in the beer. A device has been described (Chalmers <strong>and</strong><br />
Haughney, 1998) in which the shear at the membrane surface is created mechanically<br />
by vibrational energy. The basis of the design is the use of atorsional spring mass<br />
system for the creation of vibrational shear at the membrane surface (Fig. 15.12). A<br />
motor rotatesashaft with an eccentric weight. Themotion ofthiscausesthe motionof<br />
asecond weight, which has arelatively high mass moment of inertia, this is called the<br />
seismic weight. The motion ofthe seismic weight is translated through atorsion bar or<br />
spring to a membrane element assembly. The element assembly vibrates at a<br />
frequency of 50±60Hz <strong>and</strong> the energy requirement is 0.03 to 0.15kW/m 2 .<br />
The membrane assembly can contain up to 40m 2 ofmembrane area <strong>and</strong> it is made<br />
to operate three streams: feed, permeate (or filtrate), <strong>and</strong> retentate. The element<br />
assembly has atop <strong>and</strong> bottom end plate <strong>and</strong> membrane elements with spacers on the<br />
inner <strong>and</strong> outer diameter. The spacers provide an open channel for the distribution of<br />
thefeedfluid(beercontainingyeast, e.g.,tankbottoms)acrossthemembranefrom the<br />
outer diameter to the inner diameter. As this happens the filtrate (beer) passes through<br />
the membrane <strong>and</strong> is directed to acentre channel. Fluid that does not pass through the<br />
membrane (yeast) is concentrated <strong>and</strong> leaves the membrane element assembly as the<br />
retentate.<br />
Membrane quality is very important for the success of the system. Membranes have<br />
been made from ceramic tubes <strong>and</strong> hollow fibres or flat sheet polymers. Initial trial<br />
work with the vibrating membrane filter used aPTFE membrane, each element of<br />
which was 0.4 m 2 ;aten element system therefore presented 4m 2 ofseparating<br />
membrane. Beer/yeast slurries of up to 80% solids (by volume) were processed. For<br />
tank bottoms where the solids content was >60% by volume, dilution with water<br />
improved performance. Loss of carbon dioxide can be prevented by pressurizing the<br />
system<strong>and</strong>operating withabackpressureonthepermeateof1.0to1.4bar(15±20lb./<br />
in. 2 ).Yeast in lager <strong>and</strong> ale tank bottoms was concentrated to adry solids content of<br />
20% (80% by volume). Pressure drop across the membrane was not significant <strong>and</strong><br />
Knife<br />
Fig. 15.11 Rotary vacuum filter (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Divided tank<br />
(portion under<br />
knife receives<br />
solids)
Permeate port<br />
Motor mounting<br />
plate<br />
Drive shaft<br />
enclosure<br />
Motor<br />
Retenate ports<br />
Filter pack<br />
Tubing support<br />
Torsion bar<br />
Node clamp<br />
Shipping clamp<br />
Fig. 15.12 Vibrating membrane filter (Chalmers, <strong>and</strong> Haughney, 1998).<br />
successful permeate flows were obtained at pressures of 0.8 bar (12 lb./in. 2 ).<br />
Commercial systems have now been developed that can operate at average flow<br />
rates of up to 2,500 l/h which could serve a brewery producing 4,000,000 hl of beer a<br />
year of which 6% was barm (tank bottoms). Cleaning the membrane assemblies can be<br />
effected with caustic soda solutions (0.2 M at 60 ëC; 140 ëF). The process has an<br />
extremely low energy requirement of 0.004 kWh/l of recovered beer. This factor,<br />
coupled with the compactness of the machine, the ease of automation <strong>and</strong> low<br />
manpower input makes the use of a vibrating membrane filter an attractive option for<br />
separating yeast <strong>and</strong> beer in barm <strong>and</strong> tank bottoms.<br />
15.5.2 Beer filtration<br />
The final process to consider in beer treatment, prior to packaging, is filtration. This is the<br />
clarification of the beer to a st<strong>and</strong>ard that is acceptable for sale. The process involves the<br />
removal of any remaining yeast cells <strong>and</strong> the removal of precipitated protein <strong>and</strong><br />
polyphenol haze material. The beer must be rendered stable so that visible changes do not<br />
occur during its commercial (shelf) life, which could be up to 52 weeks from the date of<br />
packaging. To be successful, beer coming onto the filter must contain < 0.2 million yeast<br />
cells/ml of beer <strong>and</strong> so the processes discussed above are crucial for the success of the<br />
final beer filtration.<br />
The driving force for filtration is the pressure difference between the filter inlet <strong>and</strong><br />
outlet. Pressure is always greater at the inlet <strong>and</strong> the pressure difference is an indicator of<br />
how much the filter is resisting filtration. An increase in this pressure difference indicates<br />
the approach of the end of a filter run. In commercial brewery <strong>practice</strong> this is an<br />
important factor <strong>and</strong> long filter runs are essential to overall brewery efficiency.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Avery important factor in successful filtration is the chilling of the beer. The lower<br />
the temperature the more cold trub <strong>and</strong> chill haze will form. Filtration will remove this<br />
material, provided that the beer temperature does not rise in the filter itself. Beers<br />
emerging from maturation must therefore be maintained at 2 to 1ëC (28±30ëF)<br />
through filtration. The turbidity (see Chapter 19) of the beer leaving the filter must be<br />
< 0.5 ëEBC. In good filter <strong>practice</strong> losses of colour, extract, bitterness <strong>and</strong> foam potential<br />
should be minimal.<br />
Care must also be taken to avoid oxygen pick-up after maturation <strong>and</strong> during the<br />
filtration process. At the end of maturation the oxygen concentration should be<br />
< 0.01 mg/l. This concentration can be maintained with sound <strong>practice</strong>. Deaerated/<br />
deoxygenated water must be used for introducing powder pre-coats on powder filters <strong>and</strong><br />
carbon dioxide is used as the counter pressure gas to move beer from maturation through<br />
to the filtration process. Air must be completely removed from all pipework before the<br />
passage of beer.<br />
As with beer sedimentation there have been attempts to define filtration in<br />
mathematical terms following empirical studies. It was found in 1856 that:<br />
Q ˆ ' PA<br />
LM<br />
where Q is the flow rate in ml/second, ' is the permeability factor, P is the pressure<br />
differential in dynes/cm 2 , A is the area of the filter medium in cm 2 , M is the viscosity of<br />
the liquid in poises, <strong>and</strong> L is the thickness of the filter medium in cm. Filter throughput<br />
may therefore be increased by increasing ' which relates to the composition of the filter<br />
material. Alternatively, the applied pressure <strong>and</strong> filter area may be increased <strong>and</strong> the filter<br />
thickness <strong>and</strong> liquid viscosity decreased, this latter factor cannot be achieved in <strong>practice</strong><br />
without increasing temperature, which could result in the re-solution of haze polymers. A<br />
high viscosity is sometimes indicative of the presence in solution of high molecular<br />
weight -glucan, -glucan or yeast polysaccahride which will adversely affect filtration<br />
<strong>and</strong> if this is the case the cause should be sought in examining malt quality <strong>and</strong><br />
brewhouse performance. In considering any type of brewery filtration system the above<br />
equation should be studied to optimize performance.<br />
Different mechanisms of filtration can be used:<br />
· Sieving or surface filtration in which the particles are trapped in pores in the filter<br />
medium <strong>and</strong> retained in a layer. Filtration quality improves with time but the volume<br />
flow decreases continuously.<br />
· Depth filtration in which a separation medium, e.g., kieselguhr is used on a support<br />
<strong>and</strong> which causes the particles in the beer to take a very elongated route through a<br />
large surface area. The particles are retained by mechanical sieving because of size<br />
<strong>and</strong> will gradually block the pores in the medium <strong>and</strong> so reduce flow rate <strong>and</strong> the<br />
particles can also be retained by adsorption as a result of electrical charge effects.<br />
Traditionally, surface filtration in breweries was associated with sheet filters.<br />
However, adsorption can occur on some sheets as demonstrated when some yeasts <strong>and</strong><br />
bacteria fail to penetrate sheets when the pore size should permit it, as the fibres hold the<br />
negatively charged micro-organisms electrostatically. Breweries frequently employed<br />
double pass filtration where the beer was filtered twice through discrete systems. This<br />
could be two sets of sheet filters of decreasing pore size or a depth filter followed by a<br />
sheet filter (called a polishing filter). These systems were sometimes associated with high<br />
beer losses <strong>and</strong> the addition of oxygen to the beer <strong>and</strong> there is increasing use of single<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
pass filtration to prepare beer for packaging. Of course if sterile filtration is to be used<br />
prior to packaging then there will be aseparate filter for this purpose (Chapter 21). The<br />
two different mechanisms of filtration can be effected using sheets or by using cloths on<br />
which amedium (powder) is deposited to create the depth. Genuine surface filtration is<br />
also achieved with membranes (Chapter 21), which can be made of many polymers, e.g.,<br />
polyamide, polyethylene, <strong>and</strong> polycarbonate.<br />
Sheet filtration<br />
The st<strong>and</strong>ard filter sheet in breweries is 60 62cm <strong>and</strong> the largest normally 100<br />
100cm. Alarge single-ended filter press would have 240 filter plates allowing afiltration<br />
rate of 120hl/h. This low throughput rate has limited the use of sheet filtration for<br />
primary filtration in large breweries. Sheet filters are now usually used only as second<br />
polishing filters following depth filtration with powders or are dispensed with<br />
completely. There have been developments in the design of filter plates to ensure the<br />
maximum surface area of filter sheet is presented to the beer (Fig. 15.13). Stainless steel<br />
plates are now usually used having corrugated inserts as supports. Flow rates of > 2.0 hl/<br />
m 2 /h have been achieved.<br />
Filter sheets were originally made from a mixture of cellulose <strong>and</strong> asbestos fibres.<br />
Sheets have also incorporated kieselguhr for over 70 years. Recently perlite, glass fibres<br />
<strong>and</strong> cotton fibres have been used. Asbestos is not now used because of the carcinogenic<br />
properties of some types of asbestos. It is fair to comment that the elimination of the use<br />
of asbestos took a long time in many breweries. Asbestos was useful because it offered<br />
adsorption through electrical charge as well as surface filtration. This has been replaced<br />
in some applications by the incorporation into the sheet of aluminium oxide or zirconium<br />
oxide fibres. PVPP can also be incorporated into filter sheets to impart additional<br />
stabilization to the beer by removal of polyphenols (15.3). In large throughput systems<br />
washing the sheet with a solution of 0.5% sodium hydroxide regenerates the PVPP.<br />
Sheet filters generally operate at low flow rates. If attempts are made to increase the<br />
flow rate, pressure can force yeast <strong>and</strong> haze particles off the fibres <strong>and</strong> through the filter<br />
into the beer. For this reason pressure differentials not greater than 0.7 bar (10 lb./in. 2 )<br />
are often maintained. When this pressure differential is exceeded then sheets are cleaned<br />
by back-washing or are replaced. Sheet filters are often sterilized with steam (0.6 bar;<br />
9 lb./in. 2 ).<br />
Sheet filtration does, therefore, have a number of drawbacks. The sheets cannot be<br />
regenerated indefinitely <strong>and</strong> so operating costs are high. The filter occupies a lot of space<br />
if substantial throughput is required <strong>and</strong> it is not easy to automate. It is also very sensitive<br />
to variable solids levels in the beer <strong>and</strong> will quickly block. As a primary beer filter the<br />
sheet filter has usually been replaced by the powder filter.<br />
Powder filtration<br />
The most successful <strong>and</strong> cost effective beer clarification is achieved by powder filtration.<br />
This was developed from the earlier pulp or mass filtration in which the mechanism is<br />
primarily depth filtration. Many breweries employ this system for single pass filtration<br />
<strong>and</strong> are able to deliver beers to packaging consistently at < 0.5 ëEBC haze units. In a<br />
powder filter the powder (filter aid) is coated onto a support <strong>and</strong> provides a tortuous path<br />
through which the beer passes giving many opportunities for the trapping <strong>and</strong> adsorption<br />
of particles.<br />
Two types of powder are commonly used: kieselguhr or perlite.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(a)<br />
Beer out<br />
Beer in<br />
(b)<br />
Outlet pressure<br />
gauge<br />
Outlet sight<br />
glass <strong>and</strong><br />
sampling valve<br />
Outlet <strong>and</strong> valve<br />
Valve for draining<br />
In-plates Out-plates<br />
Filter sheet<br />
Plate Sheet Plate<br />
· Kieselguhr is a diatomaceous earth, which is mined from Miocene period deposits in<br />
Europe <strong>and</strong> North <strong>and</strong> South America. It consists of skeletons of marine algñ<br />
containing silicon dioxide. Kieselguhr powders for use in brewing are prepared by<br />
drying <strong>and</strong> milling the mined raw material. Most effective filtration was achieved with<br />
the use of calcined kieselguhr prepared by heating the raw material in rotating drums<br />
at 600 to 800 ëC (1,100±1,450 ëF). However this substance is classified as highly<br />
dangerous when inhaled <strong>and</strong> can give rise to the disease of silicosis. Equipment is<br />
needed for automatic slitting of bags <strong>and</strong> transfer to slurry tanks to avoid manual<br />
h<strong>and</strong>ling. Uncalcined kieselguhr, prepared by drying at < 400 ëC (750 ëF) represents<br />
only a moderate risk <strong>and</strong> is now usually preferred. However, some uncalcined<br />
kieselguhr can contain traces of iron <strong>and</strong> other metals. Numerous grades of kieselguhr<br />
are produced from fine through to medium <strong>and</strong> to coarse. The finer the kieselguhr the<br />
better is the clarification but the speed of filtration is less. Coarser grades give a rapid<br />
flow rate but poorer clarification. Kieselguhr usage varies from 70 to 220 g/hl. It is an<br />
(c)<br />
Inlet pressure gauge Inlet sight glass<br />
<strong>and</strong> vent valve<br />
Venting<br />
valve<br />
Inlet <strong>and</strong><br />
valve<br />
Tray Tray<br />
Drain<br />
(d) (e)<br />
Compression<br />
screw<br />
Valve for<br />
venting <strong>and</strong><br />
entry of steam<br />
Recess for<br />
compression<br />
screw<br />
Drain valves<br />
Fig. 15.13 Details of a sheet filter; (a) vertical section, (b) single plate front view, (c) alignment of<br />
plates <strong>and</strong> sheets, (d) control end of machine, (e) compression end of machine (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
expensive product <strong>and</strong> it is also expensive to dispose of as slurry, usually to l<strong>and</strong>fill<br />
sites where it is an unsatisfactory uncompacted infill.<br />
· Perlite is avolcanic material, mostly composed of aluminium silicate, obtained from<br />
Greek isl<strong>and</strong>s. Raw perlite is heated to about 750ëC (1,400ëF), which causes bursting<br />
of the particles yielding glassy structures. These are milled to afree flowing powder,<br />
which is about 30% lighter per unit volume than kieselguhr. Perlite represents alow<br />
risk to health but, because of its low density, it disperses easily in the air <strong>and</strong> creates<br />
nuisance dust. At low pH values (
(c)<br />
Direction<br />
of flow of<br />
liquor to<br />
be filtered<br />
3 Filter cake of<br />
removed impurity<br />
<strong>and</strong> filter aid<br />
particles<br />
Direction<br />
of flow of<br />
filtered<br />
liquor<br />
1 Filter cloth<br />
2 Precoat of filter<br />
aid particles<br />
Fig. 15.14 Principles of kieselguhr filtration (Hough et al., 1982).<br />
(a)<br />
(b)<br />
Filter sheet<br />
Plate Frame Plate Frame<br />
Pre-coat<br />
Bodyfeed<br />
Fig. 15.15 Plate <strong>and</strong> frame filter; (a) side view, (b) automatic mechanism for moving individual<br />
plates for opening <strong>and</strong> closing, (c) arrangement of plates <strong>and</strong> frames <strong>and</strong> the flow of beer. (Courtesy<br />
of Alfa-Laval Ltd.).<br />
into the void in the plate from which it discharges (Fig. 15.15). The sheets are<br />
washable <strong>and</strong> have a long life. After filtration the plates are separated <strong>and</strong> the<br />
kieselguhrisdislodgedfrom thesheets byspraying.Insomeplate<strong>and</strong>framefiltersthe<br />
sheet is protected by adisposable cellulosic `nappy liner', which does not add to<br />
pressure differentialbut doesprolong the life ofthe sheet<strong>and</strong> aids kieselguhrdisposal.<br />
· Leaf filters. Leaf filters (or screen filters) have aseries of stainless steel leaves fitted<br />
either vertically or horizontally inside afilter body (Fig. 15.16). In avertical leaf filter<br />
both sides of the support are coated with filter aid whereas in the horizontal leaf filter<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Beer in<br />
Beer out<br />
Vertical filter<br />
only the upper surface is used. The kieselguhr adheres to the stainless steel septa<br />
because of the pressure at which the beer is forced into the filter. Even coating of the<br />
supports is not easily achieved particularly with the vertical leaf filter in which the<br />
kieselguhr tends to slip downwards. Anew type of support for horizontal leaf filters<br />
hasbeendescribed(Oechsleetal.,2000).Thisisastainlesssteelmembraneof0.4mm<br />
gauge <strong>and</strong> an optimized slot width of 35 m<strong>and</strong> alength of 2mm. The slots widen<br />
towards the filtrate side, which reduces blocking. In trial work this support was<br />
effective in beer filtration without the use of an initial coarse kieselguhr pre-coat. This<br />
resulted in major savings in kieselguhr use as the first pre-coat can be up to 800gof<br />
kieselguhr/m 2 of filter. Time was also saved <strong>and</strong> longer filtration cycles were<br />
demonstrated with beer clarities better than the controls delivered from conventional<br />
pre-coat filtration.<br />
There is also the potential for health improvements with the elimination of the<br />
coarse kieselguhr pre-coat, which in some breweries is composed of calcined<br />
kieselguhr, which is the most risky for health. Increasingly breweries have to cope<br />
with the filtration of more different beer br<strong>and</strong>s, some in very low volumes. This is a<br />
result of greater consumer pressure <strong>and</strong> the proliferation of licence brewing where<br />
some companies will brew another company's br<strong>and</strong>s. Losses must be minimized <strong>and</strong><br />
the flexibility <strong>and</strong> automation of the filter is very important. Single pass filtration in a<br />
horizontal leaf filter is effective in achieving these aims (Thilert, 1999). Cleaning leaf<br />
filters is effected by spraying the leaves <strong>and</strong> filter body <strong>and</strong> in the case of the<br />
horizontal filters spinning the leaves to deposit the spent kieselguhr into aholding<br />
tank. Sometimes the last remains of the kieselguhr must be removed by pressure<br />
spraying from abuilt-in spray bar as aresult of which this small amount of material<br />
goes down the drain. This volume must be minimized because entrapped beer will<br />
contribute to COD of the effluent <strong>and</strong> the powder will raise suspended solids, both<br />
important factors in the charging formulñ for effluent in some countries (Chapter 3).<br />
a<br />
e<br />
Horizontal filter<br />
Fig. 15.16 Vertical <strong>and</strong> horizontal leaf filters; horizontal filter: (a) beer inlet, (b) filter support, (c)<br />
perforated shaft, (d) filtered beer outlet, (e) sediment outlet (Courtesy of Alfa-Laval Ltd.).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
b<br />
c<br />
d
(b)<br />
(a)<br />
Division plate<br />
Inlet<br />
Sludge outlet<br />
Filter head<br />
Outlet<br />
Filter body<br />
Fig. 15.17 C<strong>and</strong>le filter or Metafilter; (a) c<strong>and</strong>le filter, (b) detail of single c<strong>and</strong>le (Hough et al.,<br />
1982).<br />
· C<strong>and</strong>le filters. A c<strong>and</strong>le filter is a cylindrical, vertical pressure vessel containing<br />
many filter elements (Fig. 15.17). Each element consists of a rod of Y cross-section<br />
around which annular discs are stacked. The discs are made so that liquid can<br />
penetrate between them <strong>and</strong> then flow along the channels between the holes in the<br />
discs <strong>and</strong> the recesses of the Y-section rod. Kieselguhr powder builds up between<br />
adjacent filter discs to provide the surface area for depth filtration. Between 500 <strong>and</strong><br />
700 c<strong>and</strong>les can be arranged in a cylindrical housing to create a very large filter.<br />
Views vary considerably on the relative advantages of different types of filter. It is<br />
difficult to draw firm conclusions but some principles emerge. To avoid losses <strong>and</strong> to<br />
avoid oxygen pick-up, single pass filtration should be used wherever possible. On this<br />
basis powder filtration is usually regarded as being superior to sheet filtration. Powder<br />
filters have been compared for a number of parameters (Harding, 1977). Flow rates<br />
through filters are usually about 5 hl/m 2 but slower rates will ensure more effective<br />
particle removal. Plate <strong>and</strong> frame filters are more easily pre-coated <strong>and</strong> less susceptible to<br />
pressure surges than leaf or c<strong>and</strong>le filters but the greater volume of the filter means higher<br />
beer losses, an important factor.<br />
Sterilization of the filter, usually achieved with steam, is easier in c<strong>and</strong>le filters <strong>and</strong><br />
leaf filters than in plate <strong>and</strong> frame filters which have a large mass of metal. Kieselguhr<br />
removal is easiest with the new generation of horizontal leaf filters. (Thilert, 1999). These<br />
filters are well suited to low volume filtration for a series of different br<strong>and</strong>s with<br />
minimal losses <strong>and</strong> oxygen pick-up.<br />
It is difficult to generalize on capital costs but c<strong>and</strong>le filters are usually the cheapest<br />
<strong>and</strong> horizontal leaf filters the most expensive. The revenue cost varies considerably<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
depending on the conditions of use. Clearly any system avoiding the use of afirst precoat<br />
(Oechsle et al., 2000) may offer alower cost advantage.<br />
15.6 Special beer treatments<br />
Many beers are brewed at ahigher gravity (or alcoholic strength) than that at which they<br />
are subsequently sold. These beers are diluted after fermentation <strong>and</strong> usually after<br />
filtration just prior to packaging (Section 15.2.4). Fermentation at high gravity was<br />
considered in Chapter 12. Some beers are sold as low-alcohol or no-alcohol products <strong>and</strong><br />
the alcohol is removed after fermentation. Low-alcohol beers can also be produced by<br />
changes to the mashing process or to fermentation <strong>and</strong> all types will be discussed in this<br />
section. There has also been interest in the sale of beers produced by freezing to remove<br />
water <strong>and</strong> yield a beer with a smooth flavour <strong>and</strong> elevated alcohol content. These are<br />
known as `ice beers'. A further `special' type of beer is the so-called diet beer, which has<br />
a very low dextrin level (but often high alcohol), <strong>and</strong> although this is not strictly a postfermentation<br />
`treatment beer' it will be reviewed in this section.<br />
15.6.1 Low-alcohol <strong>and</strong> alcohol-free beers<br />
The production of alcohol-free <strong>and</strong> low-alcohol beers has a long history <strong>and</strong> patents on<br />
the processes involved go back over 100 years. Marketing <strong>and</strong> sale of these beers has<br />
varied in intensity throughout this period. Low-alcohol beers were produced in<br />
considerable volume at the time of the First <strong>and</strong> Second World Wars as a result of the<br />
shortage of raw materials <strong>and</strong> prohibition in the USA from 1919 to 1933 stimulated<br />
production. There has been renewed interest, since about 1978, because of legislation<br />
relating to the driving of motor vehicles <strong>and</strong> health considerations leading to some beliefs<br />
in the advantages of drinking less alcohol. There is also a trade in the export of nonalcoholic<br />
beer to Islamic countries where the sale of alcohol is banned. These situations<br />
lead to the development of a healthy market <strong>and</strong> most major brewers included lowalcohol<br />
<strong>and</strong> alcohol-free beers in their product portfolios. The market for these beers has<br />
recently come under pressure both from aggressive marketing from soft drink companies<br />
<strong>and</strong> from so-called `alcopops' in which alcohol is mixed with some type of fruit extract.<br />
This has led to br<strong>and</strong> losses <strong>and</strong> there are now fewer types of beer available. However the<br />
competition has resulted in marked flavour improvements in those br<strong>and</strong>s which have<br />
survived. In 1992, in Europe, the market for low-alcohol beer was 4% of the total<br />
alcoholic drinks market but it has shown no growth since this time.<br />
Legal definitions of what constitute low-alcohol <strong>and</strong> alcohol-free beer varies from<br />
country to country. A low-alcohol definition will allow an alcohol content of 0.5 to 1.2%<br />
v/v whereas an alcohol-free beer should contain < 0.5% v/v alcohol. Production methods<br />
for these beers involve either the removal of alcohol in a post-fermentation treatment or<br />
the restriction of alcohol production during the brewing process. Removal of alcohol can<br />
be effected by vacuum distillation, vacuum evaporation, dialysis <strong>and</strong> reverse osmosis.<br />
Restriction of alcohol production can involve choice of grist materials, control of<br />
mashing schedules, stopped fermentation, <strong>and</strong> use of special yeasts.<br />
Vacuum distillation<br />
In this process (Regan, 1990), the beer to be de-alcoholized is heated to 50 ëC (122 ëF) in a<br />
plate heat exchanger <strong>and</strong> is then de-esterified under high vacuum. Volatile components<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
evaporate from the beer <strong>and</strong> are collected in a mixing tank. The de-esterified beer is then<br />
separated from alcohol in a vacuum column at about 40 ëC (104 ëF) <strong>and</strong> passed to the<br />
mixing tank where it is recombined with the volatile components. This method has<br />
produced beers with sound flavours but is now seldom used because the high temperatures<br />
involved do tend to make the consistent production of high-quality beer difficult.<br />
Vacuum evaporation<br />
This process (Attenborough, 1988; Narziss et al., 1992; Regan, 1990) has developed from<br />
considering the difficulties of vacuum distillation. The temperatures used are lower than<br />
with vacuum distillation <strong>and</strong> the residence time under evaporation is less. It is easier to<br />
produce de-alcoholized beers of consistent quality with this method. A process described<br />
by Alfa-Laval involves centrifugal evaporation. Beer is pumped over the internal surface<br />
of a rotating conical heat exchanger <strong>and</strong> forms a film (0.1 mm) over this surface. The beer<br />
remains in this position for 0.5 to 1.0 seconds <strong>and</strong> reaches a temperature of 30±40 ëC (85±<br />
104 ëF). The concentrated, de-alcoholized beer is collected at the periphery of the cone<br />
where it is drawn by suction to a cooler. The aroma compounds are retained in the beer by<br />
this technique <strong>and</strong> this is the real advantage compared to the vacuum distillation method.<br />
The alcohol evaporates <strong>and</strong> the vapour passes through the centre of the cone to a<br />
condenser. The process, taking only about ten seconds, is repeated several times if it is<br />
required to lower the alcohol content to < 0.5%.<br />
An APV system utilizes a triple effect falling film evaporator. The maximum temperature<br />
of the beer is about 40 ëC (104 ëF) <strong>and</strong> the residence time in the system is three to five<br />
minutes. These plants are in wide use <strong>and</strong> can give throughputs of 200 hl/hour. Warmed beer<br />
is carefully heated in three evaporators in series <strong>and</strong> the alcohol vapour in each case is<br />
collected in a pressure reduction vessel <strong>and</strong> then condensed in a spray condenser. After this<br />
triple process the beer, at 0.3% v/v alcohol, is cooled in a heat exchanger counter-current to<br />
incoming beer, which is so warmed. The beer leaves the plant at about 1 ëC (34 ëF).<br />
Dialysis<br />
In this method (Attenborough 1988; Niefind, 1982; Regan, 1990) the alcohol is removed<br />
by pumping beer through a membrane at a pressure of just over 2 bar (30 lb./in. 2 ). The<br />
membrane is normally a hollow fibre with a very thin wall. The membranes are collected<br />
into modules of several thous<strong>and</strong> <strong>and</strong> sealed at both ends. The beer is pumped uniformly<br />
through the membranes <strong>and</strong> the dialysate (water containing alcohol) passes through the<br />
hollow fibres in the opposite direction. The rate of dialysis is directly proportional to the<br />
concentration gradients formed <strong>and</strong> inversely proportional to the size of the molecules.<br />
Equilibrium is reached when the alcohol concentration is the same on both sides of the<br />
membrane. Alcohol is therefore removed from the dialysate by continuous distillation at<br />
reduced pressure <strong>and</strong> so the process of alcohol removal can continue. Separation of beer<br />
<strong>and</strong> alcohol occurs between 1 <strong>and</strong> 6 ëC (33 <strong>and</strong> 43 ëF) <strong>and</strong> so quality of the resultant dealcoholized<br />
beer is good. However, important flavour esters pass out of the beer with the<br />
alcohol. The dialysate, therefore, passes to a rectification column for removal of alcohol.<br />
The alcohol-free fraction containing esters is reincorporated into the low-alcohol beer.<br />
This is a complex process <strong>and</strong> considerable skill is required to produce beers of consistent<br />
flavour. Nevertheless the process is in use throughout the world.<br />
Reverse osmosis<br />
This process uses filtration at high pressure (30 to 60 bar) through a semipermeable<br />
membrane. The membranes are made of cellulose acetate, nylon or other polymers <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
allow the passage of small molecules such as water <strong>and</strong> ethanol <strong>and</strong> hold back the larger<br />
molecules. The high pressure used forces the water <strong>and</strong> alcohol against the natural<br />
osmotic pressure of the beer through the membrane. The flavour <strong>and</strong> aroma compounds<br />
mostly remain in the beer although in some systems the water/alcohol mixture (permeate)<br />
is rectified <strong>and</strong> the alcohol-free fraction, which does contain some volatiles, is added back<br />
to the beer. In other systems the permeate is used without rectification for sparging.<br />
The high pressure used causes heating <strong>and</strong> the equipment must be cooled so that the<br />
beer temperature does not rise above 15 ëC (60 ëF). The membrane modules must possess<br />
a large surface area to achieve commercial flow rates of at least 25 hl of low-alcohol beer<br />
per hour. About 60 l/m 2 /h of flow is possible so to achieve the above capacity around 20<br />
modules would be needed <strong>and</strong> the plant would be expensive. Cleaning the membranes is<br />
essential for optimum performance.<br />
Control of mashing<br />
Clearly, if mashing can be performed to produce a wort of low fermentability then there<br />
is the possibility of fermenting this wort to yield a beer of low alcohol content (Muller,<br />
1990). These methods were formerly associated with intensely `worty' flavours in the<br />
resultant beers which were cloying <strong>and</strong> not `moreish'. To reduce worty flavour the<br />
amount of malt in the mash must be reduced <strong>and</strong> replaced with starchy adjunct, which can<br />
provide 40 to 70% of the extract. Mashing is carried out to restrict amylolysis, by using<br />
temperatures of 70±80 ëC (160±175 ëF), which produces high levels of non-fermentable<br />
dextrins. The pH value of the wort is also usually artificially lowered with the use of<br />
phosphoric or sulphuric acids. Used on their own, methods relying solely on the<br />
restriction of amylolysis are seldom successful for the production of non-alcohol beers<br />
<strong>and</strong> they have to be combined with further measures to control the fermentation. It is also<br />
essential to ensure vigorous wort boiling to lower the levels of aldehydes, which in the<br />
absence of normal levels of ethanol, will spoil flavour.<br />
Control of fermentation<br />
The essence of these methods is to lower ethanol production by restricting fermentation.<br />
This can be done by stopping yeast activity before fermentation is complete, by using a<br />
special strain of yeast, by temperature control, or by controlling contact time of the yeast<br />
with wort. Fermentation can be stopped early by removal of yeast by filtration or<br />
centrifugation or by using a plate heat exchanger to provide a thermal shock <strong>and</strong> so kill<br />
the yeast. These methods are difficult to control (Brenner, 1980). A flocculent yeast at<br />
low pitching rate must be used <strong>and</strong> these products require maturation for at least ten days<br />
at 1 ëC (30 ëF) to ensure an acceptable flavour.<br />
In a patented so-called `cold contact' process, yeast is mixed with wort at 0.5 ëC<br />
(31 ëF) for 48 hours <strong>and</strong> circulated to ensure good mixing (Schur, 1983). There is virtually<br />
no ethanol production but carbonyl compounds are reduced in concentration <strong>and</strong> so worty<br />
flavour is reduced. After yeast removal the product can be matured for packaging.<br />
Fermentation can also be controlled by special yeasts, e.g., Saccharomycodes ludwigii,<br />
that ferment only glucose, fructose <strong>and</strong> sucrose, which comprise about 15% of the<br />
carbohydrate in an all malt wort. The resultant beer, therefore, contains < 0.5% ethanol<br />
but tastes sweet because of the high residual maltose <strong>and</strong> maltotriose content.<br />
The problem remains with any of these restricted fermentation methods that control is<br />
difficult <strong>and</strong> the beers often have a worty taste that limits their appeal. These problems<br />
can be reduced in continuous fermentation processes with immobilized yeast cells. A<br />
successful system is in use in The Netherl<strong>and</strong>s (Mensour et al., 1997) in which the yeast<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
is immobilized on DEAE cellulose in packed beds as a result of ionic binding between<br />
the negatively charged yeast cells <strong>and</strong> the positively charged carrier. Lactic acid is added<br />
to the wort before fermentation to lower the pH value to about 4.0 <strong>and</strong> so restrict the<br />
growth of bacteria. The wort thus treated is allowed to percolate through the reactor at<br />
1 ëC (30 ëF). Under these conditions the yeast preferentially metabolizes glucose.<br />
Maltose <strong>and</strong> maltotriose are not easily consumed as a result of the repression by glucose<br />
of their transport systems. A product of < 0.1% alcohol is produced. It is low in carbonyl<br />
compounds but contains some esters associated with normal beers <strong>and</strong> has good flavour<br />
stability. The system requires cleaning <strong>and</strong> re-sterilizing twice a year. A similar system<br />
has been described (Aivasidis et al., 1991) using sintered glass beads instead of DEAE<br />
cellulose as the support medium.<br />
Use of spent grains<br />
These methods (Attenborough, 1988) utilize spent grains to produce worts of low<br />
fermentability. The grains can simply be extracted with water or by acid hydrolysis or can<br />
be extrusion cooked. Fermentation is normally at a gravity of 8 ëP (32 ëSacch). Again,<br />
these beers require long maturation times (at least 14 days) to yield acceptable flavours.<br />
It is difficult to generalize on the best method to produce low-alcohol beer of<br />
acceptable flavour that will persuade the drinker to drink more of the product. Dealcoholized<br />
beers often contain lower concentrations of potential flavour-spoiling<br />
aldehydes than those produced by restricted fermentation but they lack the higher<br />
alcohols that can contribute positively to flavour. Further development of these drinks<br />
will take place only if the market grows.<br />
15.6.2 Ice beers<br />
Brewers have long experimented with ice beer. In Germany kegs of beer were<br />
deliberately `frozen' in the winter when an ice wall would form on the inside of the<br />
container <strong>and</strong> an increasingly strong beer would result. These were dangerous beers to<br />
drink particularly as the starting point was often a bock beer of high gravity (16 ëP;<br />
64 ëSacch)!<br />
In the 1980s many brewers were interested in lowering costs as sales began to fall after<br />
the peaks of the late 1970s. One idea was to concentrate beer at the brewery by removing<br />
water <strong>and</strong> to reconstitute it at the point of sale, thus lowering distribution cost. An<br />
obvious method of removing water was to freeze the beer, thus removing pure water as<br />
ice. The beer flavour components remained in the beer in a more concentrated form.<br />
Much work was carried out but, for a number of reasons associated with capital <strong>and</strong><br />
revenue costs <strong>and</strong> the lack of a market, as a production process the method looked<br />
doomed to fail. Then, in Canada, the Labatt Company provided a whole new angle to the<br />
process. Labatt realized there was a powerful market association with the concept of `ice'<br />
beer <strong>and</strong> a new beer type was born. There was now a real market to drive product<br />
development. By the late 1990s most major brewers had produced their own ice br<strong>and</strong>s<br />
<strong>and</strong> sales increased, backed by huge advertising spends. In 1997 in the USA alone over<br />
32 million hl of ice beer was produced <strong>and</strong> volumes were increasing by 4% year on year.<br />
The rate of growth of production of these beers has slowed but nevertheless ice beer is<br />
now an important segment of the alcoholic drinks market.<br />
The higher the original gravity of the beer the lower the temperature at which the beer<br />
freezes. The important point for ice beer is that the beer does not freeze homogeneously<br />
but as the temperature falls below 0 ëC, water separates as pure ice. Some compounds,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
a<br />
d<br />
b<br />
c<br />
CO 2<br />
e<br />
NH 3<br />
NH 3<br />
NH 3<br />
NH 3<br />
insoluble at low temperatures, such as some proteins <strong>and</strong> polyphenols also separate whilst<br />
alcohol <strong>and</strong> flavour volatiles concentrate. The key to a successful commercial process is<br />
to remove the ice as it is formed so it does not remain in one place <strong>and</strong> so restrict the<br />
process (as with casual ice bock production!). This is achieved by moving the beer during<br />
the cooling stage.<br />
In the Labatt process (Fig. 15.18) fully fermented beer is cooled <strong>and</strong> then centrifuged<br />
to remove yeast. The beer is further cooled <strong>and</strong> then pumped through three heat<br />
exchangers to lower the temperature to 4 ëC (25 ëF). Small ice crystals form <strong>and</strong> the<br />
beer is then moved to the recrystallizer where the small crystals deposit on larger ones<br />
already present. The ice crystals can be filtered from the beer or removed in a<br />
hydrocyclone. A finite amount of ice is always retained in the recrystallizer. The beer of<br />
high alcohol content is held in a storage tank <strong>and</strong> adjusted to the required alcohol content<br />
with sterile, deaerated, carbonated water. The alcohol content of the processed beer is<br />
normally higher than the starting beer <strong>and</strong> the removal of polyphenols gives the beer a<br />
characteristically smooth full taste. It is this property that is appreciated by drinkers. The<br />
process is expensive <strong>and</strong> the success for the brewery depends on the market allowing the<br />
charging of a higher price for the product. Further developments in ice beers will depend<br />
on how the market develops <strong>and</strong> what prices can be sustained for the beers.<br />
15.6.3 Diet beers<br />
These are not strictly post-fermentation treatment beers but are discussed here to complete<br />
this section on special beer types. The basis of this type is a beer low in carbohydrate. A<br />
higher proportion of fermentable carbohydrate is therefore made available to the yeast than<br />
is the case in st<strong>and</strong>ard brewery fermentations. The resulting beers have a higher ethanol<br />
content but lower dextrin levels from a given original extract compared to normal beers. It<br />
should be noted that these beers seldom have lower calorific values than normal beers,<br />
merely lower carbohydrate contents. It follows that these beers are usually derived from<br />
fermentations of 100% apparent attenuation. Mashing can be adjusted by extending the<br />
NH 3<br />
f f f<br />
NH 3 NH 3 NH 3<br />
Fig. 15.18 Ice beer plant, Labatt/Niro; (a) route from fermentation, (b) dropping cooler, (c)<br />
centrifuge, (d) yeast, (e) beer cooler, (f) heat exchanger, (g) recrystallizer, (h) route to conditioning.<br />
(Kunze, 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
h<br />
CO 2<br />
g
time <strong>and</strong> by a low temperature st<strong>and</strong> at 50 ëC (122 ëF) for at least 30 minutes but attenuation<br />
limits of more than 90% are seldom produced. Enzymes must therefore be added to the<br />
fermenter to degrade residual dextrins during fermentation <strong>and</strong> this can be in the form of<br />
malt flour or diastatic malt extract. The -amylase, -amylase, <strong>and</strong> limit dextrinase so<br />
added results in the degradation of dextrin in the fermenting wort.<br />
In some countries the addition of enzymes of fungal origin is permitted (glucoamylase<br />
<strong>and</strong> pullulanase). Beers produced using fungal enzymes tend to be more biologically <strong>and</strong><br />
non-biologically stable. Fermentation of these highly fermentable worts yields beers of<br />
very high alcohol contents <strong>and</strong> virtually no residual carbohydrate. The beers can be<br />
diluted to the appropriate alcohol content for sale. Pasteurization must be effectively<br />
carried out if using malt enzymes, as almost certainly bacteria will be introduced into the<br />
wort from the malt flour.<br />
Work at <strong>Brewing</strong> Research International in the UK in the mid 1990s (Baxter, 1995)<br />
resulted in the genetic modification of a brewing yeast strain to include a glucoamylase<br />
gene from a non-brewing yeast. This strain was approved for use <strong>and</strong> is used in-house for<br />
the production of low carbohydrate beers. But, as the result of general public disquiet<br />
about the use of genetic manipulation, there have been no commercial developments<br />
using this strain. There seems to be a continuing, if static, market for this type of beer<br />
particularly as it is often assumed (wrongly!) to be a less fattening beer.<br />
Finally, the term `light' beer is sometimes confused with diet beer. Light beers have no<br />
real generic definition but are merely beers of low-alcohol or low-sugar content. As a<br />
result what constitutes a light beer can vary enormously. Beers described as light beers<br />
can have alcoholic strengths of from 2.5 to 4.0% abv but usually have a dextrin content of<br />
about 1% <strong>and</strong> a calorific value of 25 to 30 kcal/100 ml. As such, if drunk as an alternative<br />
to normal beer, they may be less fattening.<br />
15.7 Summary<br />
The post-fermentation treatment of beer is critical in yielding a product that is fit for sale<br />
<strong>and</strong> meets the consumer's expectation of a quality drink. The aim has to be to delight the<br />
consumer so that he will return to the product <strong>and</strong> drink it again. Flavour, clarity <strong>and</strong><br />
stability of beer are improved by the post-fermentation treatments that have been<br />
discussed. This is also the area of the whole brewing process where emphasis can be<br />
placed on the development of special beer types.<br />
There is increasing interest in the manipulation of beer properties post-fermentation to<br />
produce different beer br<strong>and</strong>s from essentially the same brewhouse <strong>and</strong> fermentation<br />
techniques. This can result in considerable saving in capital <strong>and</strong> revenue cost with fewer<br />
requirements to invest in expensive maturation storage. In this respect further<br />
development in immobilized yeast technology with its potential for accelerated flavour<br />
improvement <strong>and</strong> production of novel beers would seem to be worthwhile.<br />
15.8 References<br />
AIVASIDIS, A., MANDREY, C., ELLIS, H. G. <strong>and</strong> KATZKE, M. (1991) Proc. 23rd Congr Eur. Brew. Conv,<br />
Lisbon, 569.<br />
ANDERSON, R. G., BRITES SANCHES, A., DEVREUX, A., DUE, J., HAMMOND, J., MARTIN, P. A., OLIVER-DAMEN,<br />
B. <strong>and</strong> SMITH, I. B. (2000) Fermentation <strong>and</strong> Maturation Manual of Good Practice, European<br />
Brewery Convention, Zoeterwoude, The Netherl<strong>and</strong>s.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ANDREWS, J. M. H. (1997) Ferment, 10, 309.<br />
ANNESS, B. J. (1980) J. Inst. <strong>Brewing</strong>, 86, 134.<br />
ATTENBOROUGH, M. W. (1988) Ferment, 1 (2), 40.<br />
BARNES, Z. C. (2001) Personal communication.<br />
BAXTER, E. D. (1995) Ferment, 8, 307.<br />
BRENNER, M. W. (1980) Tech. Quart. MBAA. 17, 185.<br />
BROWN, D. G. W., CLAPPERTON, J. . F. <strong>and</strong> MEILGARD, M. C. (1978) J. Amer. Soc. <strong>Brewing</strong> Chemists, 36, 73.<br />
BYRNE, H., MATHEWS, S., MADIGAN, D., KELLY, R. J., MCENROE, C. <strong>and</strong> HARMEY, D. (1999) Proc. 7th Conv.<br />
Inst. Brew (Africa Section), Nairobi, 55.<br />
CHALMERS, S. <strong>and</strong> HAUGHNEY, H. ( 1998) Proc. 25th Conv. Inst. Brew (Asia Pacific Section), Perth, 165.<br />
CHAPON, L. (1994) Brewers' Guard. 123 (12), 21.<br />
COORS, J. H. (1977) The Practical Brewer, Master Brewers' Association of the Americas. Madison,<br />
Wisconsin.<br />
DICKENSON, C. J. <strong>and</strong> ANDERSON, R. G. (1981) Proc. 18th Congr. Eur. Brew. Conv., Copenhagen, 413.<br />
ESNAULT, E. (1995) Brewers' Guard., 124 (1), 25.<br />
GARDNER, D. J. S. (1993) Ferment, 6, 279.<br />
GOPAL, C., REHMANJI, M., MOLA, A., NARAYANAN, K., TRINH, T. <strong>and</strong> WHITTINGHAM, J. (1999) Proc. 7th<br />
Conv. Inst. Brew. (Africa Section), Nairobi, 62.<br />
GOPAL, C. <strong>and</strong> REHMANJI, M. (2000) Brewers' Guard., 129 (5).<br />
GREEN, H., SHAW, R. <strong>and</strong> CANDY, E. (2000) The Brewer, 86, 201.<br />
GUZMAN, J. E., MCKEOWN, I. . P., GLEAVES, M., STEWART, G. G. <strong>and</strong> DOYLE, A. (1999) Tech. Quart. MBAA.<br />
36, 227.<br />
HANNEMAN, W. (1999) Tech. Quart. MBAA. 36, 167.<br />
HARDING, J. A. A. (1977) Brewers' Guard., 106 (9), 90.<br />
HASIMOTO, N. (1973) Ann. Rept. Res. Labs. Kirin Brew. Co., 16, 1.<br />
HOUGH, J. S., BRIGGS, D. E., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1982) Malting <strong>and</strong> <strong>Brewing</strong> Science Volume 2,<br />
2nd Edition, Aspen, Gaithersburg, Maryl<strong>and</strong>.<br />
JENDE-STRID, B. (1997) Proc. 25th Congr. Eur. Brew. Conv, Maastricht, 101.<br />
JEPSEN, S. (1991) 4th Meeting on the Industrial Applications of Enzymes, Barcelona.<br />
KUNZE, W. (1999) Technology <strong>Brewing</strong> <strong>and</strong> Malting, Int. Edition, (Translated by Wainwright, T.) VLB,<br />
Berlin.<br />
LEACH, A. A. (1989) Ferment, 2, 33.<br />
LEACH, A. A. <strong>and</strong> BARRETT, J. E. (1967) J. Inst. <strong>Brewing</strong>, 73, 246.<br />
LONG, D. E. (1995) Ferment, 8, 239.<br />
MARRIOTT, R. (1999) Proc. 7th Conv. Inst. Brew. (Africa Section), Nairobi, 140.<br />
MCKEOWN, I. P. <strong>and</strong> EARL, G. J. (2000) Brewers' Guard., 129 (6).<br />
MCMURROUGH, I., MADIGAN, D., KELLY, R. <strong>and</strong> O'ROURKE, T. (1999) Food Technology, 53, 58.<br />
MENSOUR, M. A., MARGARITIS, A., BRIENS, C. L., PILKINGTON, H. <strong>and</strong> RUSSELL, I. (1997) J. Inst. <strong>Brewing</strong>,<br />
103, 363.<br />
MIEDANER, H. (1978) The Brewer, 64 (2), 33.<br />
MULLER, R. (1990) Ferment, 3 (4), 224.<br />
MUSCHE, R. A. <strong>and</strong> DE PAUWE, C. (1999) J. Inst. <strong>Brewing</strong>, 105, 386.<br />
NARZISS, L., WULFINGER, H., STICH, S. <strong>and</strong> LAIBLE, R. (1992) Brauwelt, 51/52, 2650.<br />
NIEFIND, H. (1982) Monatsshrift fuÈr Brauerei, 35, 90.<br />
OECHLSE, D., ASCHER, R. <strong>and</strong> FEIFEL, K. (2000) Tech. Quart. MBAA., 37, 377.<br />
O'ROURKE, T. (2000) Brewers' Guard., 129 (2), 29.<br />
PAJUNEN, E. <strong>and</strong> JAÈ AÈ SKELAÈ INEN, K. (1993) Proc. 24th Congr. Eur. Brew. Conv. Oslo, 559.<br />
PAJUNEN, E., GROÈ NQUIST, A. <strong>and</strong> RANTA, B. (1991) Proc. 23rd Congr. Eur. Brew. Conv., Lisbon, 361.<br />
PFISTERER, E. <strong>and</strong> STEWART, G. G. (1975) Proc. 15th Congr. Eur. Brew. Conv., Nice, 255.<br />
REGAN, J. (1990) Ferment, 3 (4), 235.<br />
SHIEL,P. (1999) Proc. 7th Conv. Inst. Brew. Conv (Africa Section), Nairobi, 76.<br />
SCHUR, F. (1983) Proc. 19th Congr. Eur. Brew. Conv., London, 353.<br />
SIEBERT, K. J., TROUKHANOVA, N. V. <strong>and</strong> LYNN, P. Y. (1996) J. Agric. Food Chem., 44, 80.<br />
SOLE, S. M. (2000) Ferment, 13 (4), 25.<br />
STEWART, G. G., BRYCE, J. H., COOPER, D., MONAGAS, M. <strong>and</strong> YOUNIS, O. (1999) Proc. 7th Conv. Inst. Brew<br />
(Africa Section), Nairobi, 100.<br />
THILERT, T. (1999), Tech. Quart. MBAA. 36, 427.<br />
ZANGRANDO, T. <strong>and</strong> GIRINI, G. (1969) Proc. 12th Congr. Eur. Brew. Conv., Interlaken, 445.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
16<br />
Native African beers<br />
16.1 Introduction<br />
African beers almost certainly have ancient origins, <strong>and</strong> may have originated in Egypt or<br />
Mesopotamia, where beers were being made by at least 3500 BC, <strong>and</strong> probably much<br />
earlier (Briggs, 1998). The names given to African beers are often unacceptable or<br />
inexact. Thus beers brewed in southern Africa have been called Kaffir or Bantu beers (but<br />
elsewhere in Africa beers are made by peoples who are not Bantu). Another term is<br />
opaque beers but not all African beers are truly opaque (e.g. Nigerian `otika') <strong>and</strong>, on the<br />
other h<strong>and</strong>, some European-style beers are at least turbid (wheat beers; some ales<br />
consumed with conditioning yeast in suspension) <strong>and</strong> others, such as stouts, are not<br />
transparent. The term `sorghum beers' is inexact as sorghum (raw grain or malt) is<br />
sometimes wholly or largely replaced with maize or millets or wheat or barley, <strong>and</strong><br />
indeed in some cases bananas or manioc (cassava) serve as starchy adjuncts. Bouza may<br />
be made with wheat, barley or millets. Native names are also confusing in that different<br />
names are used for similar products by different tribes <strong>and</strong> within one language group<br />
different names are used for different types or qualities of beers (Daiber <strong>and</strong> Taylor,<br />
1995; Dendy, 1995; Haggblade <strong>and</strong> Holzapfel, 1989; Harris, 1997; Miracle, 1965;<br />
Novellie, 1966, 1968, 1977; Novellie <strong>and</strong> De Schaepdrijver, 1986; Peterson <strong>and</strong> Tressler,<br />
1965; Schwarz, 1956). Among the best known of these names are utshwala (Zulu) <strong>and</strong><br />
joala (Basuto).<br />
Traditionally, beers are made by women brewsters, as was the case in mediaeval<br />
Europe, <strong>and</strong> they may be consumed with some ceremony. However, in southern Africa,<br />
as urbanization occurred, men moved into towns as casual labour, leaving their<br />
womenfolk behind. To meet the dem<strong>and</strong> for opaque beers commercial brewing began<br />
(around 1908±1910) in Bulawayo <strong>and</strong> Durban. Since then, in some periods, the rate of<br />
increase in production has risen at an astonishing rate. For example, in South Africa, at<br />
one stage production increased by 26% in one year <strong>and</strong> while in 1953/4 production was<br />
20 million imp. gallons (0.909 million hl) in 1965/6 production was 120 million imp.<br />
gallons (5.46 million hl; Novellie, 1968). At the same time much larger volumes of beer<br />
were produced in homes <strong>and</strong> small-scale `village' breweries. Estimates of more recent<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
opaque beer production in different countries are in millions of hectolitres/year (Harris,<br />
1997). At first the industrialized production of these beers was not straightforward, <strong>and</strong><br />
many difficulties were encountered. To overcome these problems the CSIR (the Council<br />
for Scientific <strong>and</strong> Industrial Research) in South Africa established, around 1953, a<br />
research organization to investigate the bases of malting sorghum <strong>and</strong> beer production.<br />
The publications from this group, <strong>and</strong> its successors, provide nearly all the available<br />
information on the <strong>science</strong> of opaque beer production. Industrial production is spreading<br />
into other countries (Harris, 1997). In many areas the traditional brewing methods differ<br />
significantly from those used in southern Africa, <strong>and</strong> no doubt many brewing methods<br />
have not been described. <strong>Brewing</strong> may occur in the home, for home consumption or for<br />
sale, or it may be produced in a factory. Factory brewing seems to be carried out by men,<br />
a change that parallels the historical move from home-based brewsters (women) to<br />
industrial brewers (men) in Europe.<br />
16.1.1 An outline of the stages of production<br />
Stages usually distinguished in the production of African beers are the selection of the<br />
raw materials, malting (usually sorghum or millets or, less usually, maize), grinding,<br />
souring, cooking (with adjuncts), mashing (or conversion), straining, fermentation, <strong>and</strong><br />
packing, distribution <strong>and</strong> consumption. The souring fermentation stage produces a<br />
desired level of acidity, caused by lactic acid. The second fermentation produces (mainly)<br />
alcohol <strong>and</strong> carbon dioxide. The beers are consumed while they are warm <strong>and</strong> are still<br />
fermenting <strong>and</strong> effervescent <strong>and</strong> so their composition is continually altering. Because of<br />
the continuing production of carbon dioxide the beers are held in vented containers. If not<br />
consumed in a day or two they become flat, too acidic <strong>and</strong> poorly flavoured, at least<br />
partly because acetic acid accumulates, <strong>and</strong> they are rejected.<br />
The short shelf-lives of African beers create many commercial production <strong>and</strong><br />
distribution problems. There are major differences between African- <strong>and</strong> European-style<br />
beers. African beers are rarely or never flavoured by herbs (in contrast to hopped<br />
European-style beers), complete starch conversion is avoided, <strong>and</strong> brewing does not<br />
produce an excess of yeast. Indeed factory brewing is a net consumer of yeast. Then the<br />
beers are always consumed, with the yeast they contain, while still fermenting <strong>and</strong> they<br />
are mostly opaque because of suspended yeast, starch granules <strong>and</strong> small particles of<br />
cereal grains, which are maintained in suspension by the rising bubbles of carbon dioxide<br />
<strong>and</strong> the high viscosity of the beer. The high viscosity is caused by gelatinized but<br />
incompletely degraded starch. It seems that most drinkers are more concerned with the<br />
flavour <strong>and</strong> body of a beer than with the (changing) alcohol content. South African beers<br />
are described as being as refreshingly sour as yoghurt, with a characteristic fruity odour.<br />
Alcohol contents of 1±8% have been noted, but values of 2.5±4.5% seem to be usual.<br />
Colours vary from a pale buff to a pinkish-brown, or elsewhere may even have a reddish<br />
tinge. pH values may be 3.3±3.6, lactic acid contents about 0.26% <strong>and</strong> total solids are<br />
around 6%. However, there are wide variations in beer composition, particularly in homebrewed<br />
beers (Haggblade <strong>and</strong> Holzapfel, 1989; Harris, 1997; Novellie, 1966; Novellie<br />
<strong>and</strong> De Schaepdrijver, 1986).<br />
16.1.2 Bouza<br />
Bouza (bouzah, bowza, etc.) is a bread-beer, made in Egypt <strong>and</strong> the Sudan from wheat,<br />
barley or millets, using methods supposedly resembling those used by the ancient<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Mesopotamians <strong>and</strong> Egyptians (Briggs, 1998; Morcos et al., 1973). Coarsely ground<br />
grain, sometimes mixed with a little malt, is mixed with water <strong>and</strong> some leaven or some<br />
yeast in sourdough. After st<strong>and</strong>ing the dough is moulded into loaves, which are lightly<br />
cooked. About one-third more grain is malted <strong>and</strong> then, often after drying in the sun, the<br />
malt (green or dry) is ground up with water <strong>and</strong> lumps of the broken up loaves. The<br />
mixture begins to ferment, either spontaneously or after the addition of some older bouza.<br />
After a period of active fermentation the mixture is filtered, for example through a<br />
horsehair sieve. The introduction of air at this stage checks the fermentation, but this is<br />
soon resumed. The drink is thick <strong>and</strong> yeasty, pale yellow, acidic <strong>and</strong> with a characteristic<br />
odour. The pH may be 3.5±4 <strong>and</strong> the alcohol content 4±5.5 g/100g. It must be consumed<br />
quickly, before deterioration begins (Briggs, 1998).<br />
16.1.3 Merissa<br />
In the Sudan a beer called merissa is produced. It has been said that sometimes the women<br />
chew some of the grain <strong>and</strong> spit the mix into the mixture, so adding salivary -amylase,<br />
which may accelerate starch degradation; others do not mention this <strong>practice</strong>. Dirar (1978)<br />
describes a complex scheme for making merissa. Some sorghum grain is malted, dried <strong>and</strong><br />
ground to a flour. Raw sorghum grain is ground to a fine flour, which is divided into three<br />
equal lots, each of which is processed differently. The first lot is lightly cooked, `halfcooked',<br />
to a grey powder. The second lot is well cooked to give a brown paste. These two<br />
solid materials are mixed on leaves <strong>and</strong> allowed to cool. The third portion is wetted with just<br />
enough water to moisten it, <strong>and</strong> is set aside for about 36 hours, when a spontaneous, mainly<br />
lactic fermentation occurs. The acidified dough is strongly cooked in a steel container with<br />
repeated mixing until it is dark brown, is extremely sour <strong>and</strong> has a pleasant caramel flavour.<br />
It is cooled to room temperature <strong>and</strong> mixed with about 5% malt flour, water <strong>and</strong> some good<br />
merissa. Fermentation is well established after 4±5 hours. This material is too acid to drink.<br />
Portions of the combined two-thirds cooked flour, mixed with malt flour, are added in<br />
increments to the, strongly fermenting, acid fraction, without stirring them together. After<br />
8±10 hours. fermentation the mixture is filtered through cloth. The liquid is consumed while<br />
it is still fermenting. The solids removed by straining are fed to cattle. The product has a pH<br />
of about 4, <strong>and</strong> an alcohol content of about 5%.<br />
16.1.4 Busaa <strong>and</strong> some other beers<br />
Busaa, <strong>and</strong> similar drinks, are made in Kenya, Ug<strong>and</strong>a <strong>and</strong> Tanzania ((Nout, 1980;<br />
O'Rourke, 2001). In making busaa maize grits are mixed with water <strong>and</strong> allowed to st<strong>and</strong><br />
for 2±3 days, at about 25 ëC (77 ëF), for souring. Malted finger millet (Eleusine coracana)<br />
is prepared by steeping for 8±24 h, then germinating for 2±3 days, also at about 25 ëC<br />
(77 ëF), followed by drying in the sun for 1±2 days <strong>and</strong> coarsely grinding. The soured<br />
maize dough is cooked on steel sheets over a charcoal fire at 65±75 ëC (149±167 ëF) for<br />
three hours. The cooled maize soured material is broken into lumps <strong>and</strong> one part is mixed<br />
with 1.5 parts of water <strong>and</strong> 0.1 part of malt flour. The main fermentation proceeds for 2±4<br />
days. The mixture is then strained <strong>and</strong> consumed within one day, as it deteriorates rapidly<br />
<strong>and</strong> the increasing acidity (sometimes approaching 2% lactic acid) is unpleasant <strong>and</strong><br />
causes the suspended solids to precipitate. When consumed beers may contain 0.5±1%<br />
lactic acid <strong>and</strong> 2±4% (v/v) alcohol.<br />
Investigations into this process indicate that temperature control <strong>and</strong> the use of pure<br />
cultures of Lactobacilli <strong>and</strong> yeast could be used to advantage in the souring <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
fermentation stages, giving more stable products that could be kept longer if preserved by<br />
end-fermentation or pasteurization in bottle. The latter was preferred. Beers made by<br />
related processes include ajou (Ug<strong>and</strong>a) <strong>and</strong> mbweje (Tanzania) (O'Rourke, 2001). In the<br />
former a paste of ground millet <strong>and</strong> water is soured by burying bags of it in the ground for<br />
5±7 days. In the latter bananas are used as starchy adjuncts. In the foregoing examples<br />
some materials are well-cooked, giving the possibility of adjusting the flavours of the<br />
beers by varying the intensity of the cooking. This is not so in subsequent examples,<br />
where water-grain mixtures are only boiled.<br />
The preparation methods of other African beers have been described, otika from<br />
sorghum (Ogundiwin, 1977), pito from maize or sorghum (Ekundayo, 1969), oyokpo<br />
from millet (Pennisetum typhoideum; Iwuagwu <strong>and</strong> Izuagbe, 1985) <strong>and</strong> burukutu, also<br />
from sorghum (Faparusi, 1970; Faparusi et al., 1973) all in Nigeria. The preparation<br />
methods for sorghum beers in the Cameroons <strong>and</strong> Togo have also been described<br />
(Chevassus-Agnes et al., 1976; Perisse et al., 1959). However, most information is<br />
available for the brewing methods used in southern Africa.<br />
16.1.5 Southern African beers<br />
A reliable method of brewing, in use by the Zulus in 1907 (quoted by Fox, 1938) used<br />
sorghum for preference, but millets <strong>and</strong>/or maize might also be used. The process required<br />
considerable skill. Grain sewn into sacks was steeped in running water for 1±2 days<br />
(sorghum) or up to 4 days (maize), longer periods being used when the weather was cool.<br />
Grain was sprouted, still in the sack or in a pot, for 2±5 days, maize taking longer, until<br />
shoot growth was judged adequate (1.9 cm, (0.75 in.), for sorghum, 1.3 cm, (0.5 in.), for<br />
maize). Usually the malt was dried in the sun or inside a hut, but sometimes it was used<br />
undried. Initially unground grain or (better) a 50:50 mixture of grain <strong>and</strong> malt was soaked<br />
in water for a day. After draining the wet grain was finely ground to a paste, between<br />
stones, in the morning <strong>and</strong> the dough was moulded into lumps. In the afternoon the paste<br />
was just covered with boiling water in a pot <strong>and</strong> cold water was added to adjust the<br />
temperature according to the brewster's judgement. As the mixture slowly cooled so<br />
spontaneous acidification occurred. Next day the water was collected from above the<br />
dough <strong>and</strong> was boiled with more water while the dough itself was mixed with fresh boiling<br />
water <strong>and</strong> was mixed to a thin porridge which was added to the boiling water. After the<br />
boil, of 20±40 minutes, (longer periods being needed for maize), the mixture had<br />
thickened because the starch present had gelatinized. The mixture was too thick to pour<br />
from a spoon. Most of the mixture was allowed to cool quite slowly, but a small amount<br />
was cooled quickly <strong>and</strong> was mixed with ground malt, when starch conversion <strong>and</strong> a<br />
spontaneous fermentation began, creating a `starter culture'. When the main mash was<br />
cool enough more ground malt, in an amount exceeding the initial amount of grain by<br />
about 25%, was mixed in, together with the fermenting `starter culture', initiating a rapid<br />
onset of fermentation. When fermentation was vigorous the mixture was strained through<br />
a woven grass strainer. The filtered liquid continued to ferment while the strainings were<br />
reserved for making a `small beer'. Fermentation went on for 1±2 more days before the<br />
beer was consumed. Beyond this period the product spoilt. Analyses of these beers gave<br />
estimates of solids contents of 5±13%, alcohol contents of 0.5±8.0% (v/v; usually 4%<br />
when fresh), crude protein contents of 0.7±1% <strong>and</strong> mineral salts contents of 0.18±0.36%.<br />
Others reported solids contents of up to 20%. These beers were regarded as foods, which<br />
constituted most or all of men's diets in some seasons, <strong>and</strong> were an alternative to porridges<br />
or acidified porridges which were made without the second, alcoholic fermentation.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Other, less complicated methods have been quoted (Haggeblade <strong>and</strong> Holzapfel, 1989;<br />
Novellie <strong>and</strong> De Schaepdrijver, 1986). For example, maize broken up by pounding is<br />
mixed with boiling water <strong>and</strong> is left for a day, when a spontaneous lactic fermentation<br />
occurs. More water is added, the mixture is boiled for 2±3 h, <strong>and</strong> is left to cool. When<br />
cool a roughly equal amount of pounded sorghum malt is mixed into the soured maize<br />
adjunct mixture. After a fermentation period of 24 h the mixture is strained <strong>and</strong> is ready<br />
for consumption. In a `generalized' scheme for home or small-scale brewing for sale<br />
malted sorghum, or less usually malted pearl millet (Pennisetum typhoides) or finger<br />
millet (Eleusine coracana) is broken up using a mortar <strong>and</strong> pestle, or by grinding between<br />
stones or is pulverized in a hammer mill. In southern Africa commercially prepared <strong>and</strong><br />
ground sorghum malt may be purchased. For souring some of the malt is mixed with<br />
water <strong>and</strong> is heated for 30±90 minutes. Heating may be in traditional clay pots, or in iron<br />
or steel containers. The mixture is allowed to cool overnight, when a spontaneous lactic<br />
acid fermentation begins, <strong>and</strong> the mixture is soured. Next day the sour is mixed with more<br />
water <strong>and</strong> ground sorghum, sorghum malt, maize or maize grits. The mixture is cooked<br />
by bringing it to the boil <strong>and</strong> boiling for 2±7 h. The mixture is cooled overnight,<br />
sometimes by dividing it between several shallow dishes. During this time the mixture<br />
thickens as the gelatinized starch tends to set. In the morning mashing <strong>and</strong> fermentation<br />
begin when more malt flour is mixed in, sometimes together with some good beer, which<br />
provides an inoculum of yeasts <strong>and</strong> other microbes. This stage may be carried out in<br />
wooden, metal, clay or plastic containers. After about two days fermentation the beer is<br />
strained to remove coarse particles by passing it through a woven-grass, bag-like<br />
container (which may be squeezed to recover more liquid) or a metal screen. When<br />
fermentation has resumed the beer is ready for consumption.<br />
16.2 Malting sorghum <strong>and</strong> millets<br />
In southern Africa sorghum malts are preferred but elsewhere in Africa malts may also be<br />
made from various millets (Briggs, 1998; Haggblade <strong>and</strong> Holzapfel, 1989; Miracle, 1965;<br />
Nout <strong>and</strong> Davies, 1982; Novellie, 1968; Novellie <strong>and</strong> De Schaepdrijver, 1986; Nzalibe<br />
<strong>and</strong> Nwasike, 1995). All these plants belong to the grass family. Technically the grains<br />
are caryopses, fruits in which the ovary wall remains investing the seed as the pericarp. It<br />
seems that maize is malted only as a last resort. In `home' brewing the chosen grain is<br />
steeped sewn into rush bags, sacks or held in baskets, in running water, or in pots, jars,<br />
tubs, gourds, calabashes or other vessels, in still water. Steeping times vary from 1±3<br />
days. The grain is drained <strong>and</strong> placed in jars or baskets lined with leaves or on mats, <strong>and</strong><br />
is covered with leaves <strong>and</strong> left to germinate. From time to time it is watered. When<br />
growth is far enough advanced, in 2±6 days, the grain is usually dried in the sun before<br />
use, although sometimes the malt is used undried. It is used with the roots <strong>and</strong> shoots still<br />
attached. All the grain species occur in many varieties of widely differing malting<br />
qualities <strong>and</strong> characters. Sorghums with grains having thous<strong>and</strong> corn dry weights (TCW)<br />
of 7±61 g are known, but values of 10±38 g are more usual, these grains having<br />
dimensions of 3±5 mm (0.118±0.197 in.) by 2±5 mm (0.079±0.197 in.). Sorghum varieties<br />
vary greatly in their malting qualities.<br />
Maize grains are larger, for example dent corn may have TCWs of 150±300 g <strong>and</strong> be<br />
12 mm by 8 mm by 4 mm (0.472 by 0.315 by 0.157 in.). This grain was introduced into<br />
Africa, perhaps in the 16th century, <strong>and</strong> grains, or materials made from them, are<br />
common brewing adjuncts. The various millets have much smaller grains. This creates<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
problems for `industrial' maltsters. Like sorghum, millets have many different names<br />
(Briggs, 1998). Pearl millet (Pennisetum typhoides) has the largest grains (TCW 5±10 g;<br />
3±4 mm, 0.118±0.157 in., length). Finger millet (Eleusine coracana) is frequently malted.<br />
Common millet (Panicum miliaceum) grains are about 3mm (0.118 in.) long <strong>and</strong> have a<br />
TCW of about 6 g. Acha, or fonio (Digitaria exilis) grains have a TCW of only 0.65 g.<br />
In southern Africa, the first industrial malting of sorghum followed village <strong>practice</strong>s<br />
<strong>and</strong> to some extent this is still the case (Daiber <strong>and</strong> Taylor, 1995; Haggblade <strong>and</strong><br />
Holzapfel, 1989). Grain is steeped for about 16±18 h in concrete tanks, metal drums or<br />
barrels, <strong>and</strong> then, after draining, is spread out, in the open, on slightly sloping concrete<br />
floors in beds 13±90 cm (5.12±35.4 in.) thick. Here it is covered with wet sacks. At<br />
intervals the grain is unevenly wetted by hosing, <strong>and</strong> it may be turned by h<strong>and</strong>. In warm<br />
weather the grain is spread more thinly <strong>and</strong> in cold weather the bed is thickened to favour<br />
heat loss <strong>and</strong> heat retention respectively. Evidently temperature control is inadequate, as<br />
is regulation of the wetting, `sprinkling'. Growth is very irregular, the grain at the top of<br />
the bed being poorly grown, while that at the base, on the floor, is overgrown. When<br />
growth is judged to be sufficient the sacks are removed <strong>and</strong> the grain is spread more<br />
thinly to dry in the sun. This process is used to make malt that is sold after grinding <strong>and</strong><br />
makes a very irregular product. It often carries a high load of microbes. Apart from<br />
sometimes covering the floors with roofs, but with no side walls, <strong>and</strong> sometimes using<br />
steeps containing formaldehyde (see below) this form of malting seems to have advanced<br />
very little.<br />
Malting for the larger breweries is carried out indoors, under more controlled <strong>and</strong><br />
hygienic conditions. The grain is thoroughly cleaned before use <strong>and</strong> may be washed.<br />
Malting plant is regularly cleaned. Steeping, for 16±24 h (or even as little as 6 h), is<br />
usually carried out in tanks that may be aerated. The moisture content finally achieved<br />
is about 35%. To control microbes the grain may initially be steeped in a solution of<br />
formaldehyde or sodium hypochlorite (an agent that taints barley malts, giving them an<br />
`antiseptic' flavour). The formaldehyde treatment was originally adopted to deal with<br />
high-tannin, birdproof sorghums that are so rich in tannins they inactivate <strong>and</strong><br />
insolubilize the malt enzymes during mashing <strong>and</strong> inhibit the souring process, so<br />
blocking brewing. For the first four hours of steeping, the grain is immersed in a<br />
solution of formaldehyde (0.02±0.08%, depending on the tannin content of the<br />
particular grain), then it is thoroughly rinsed <strong>and</strong> steeping is completed in fresh water<br />
(Daiber, 1975; 1978). Recent studies indicate that steeping for longer periods of up to<br />
40 h at 25±30 ëC (77±86 ëF), with air-rests, gives superior malts (Dewar et al.,<br />
1997a,b). Evidently there would be advantages to using temperature-controlled steeps,<br />
equipped for carbon dioxide extraction, during air-rest periods.<br />
Germination is carried out in modified Saladin boxes. The grain bed, which may be<br />
1.5 m (4.92 ft.) deep, rests on a perforated base through which temperature-controlled <strong>and</strong><br />
humidified air can be blown to cool the grain. Because the grain grows so vigorously the<br />
airflow must be larger than that used for barley, around 1,000±1,200 m 3 /h/tonne grain.<br />
Because of the extensive embryo growth the volume of the grain bed increases greatly<br />
during germination. The grain should be at 24±30 ëC (75.2±86 ëF). Turning may be<br />
mechanical but, because the seedlings are so easily damaged, it may be carried out better<br />
by h<strong>and</strong>, the grain being shovelled from one compartment to the next. During turning the<br />
grain may be also sprinkled with water as required. After 5±7 days, or 4±6 days under<br />
ideal conditions, the malt is dried by blowing warm air, at 50±60 ëC (122±140 ëF), up<br />
through the grain. The temperature is kept low to minimize the destruction of enzymes,<br />
<strong>and</strong> avoiding transfer to a kiln avoids damage to the green malt. Experience with barley<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
suggests that a lower initial drying-air temperature (40 ëC; 104 ëF) would be beneficial for<br />
enzyme survival.<br />
The main objectives of malting sorghum are to generate the enzymes needed for<br />
mashing <strong>and</strong> to supply sufficient soluble nitrogen to support the Lactobacilli during<br />
souring <strong>and</strong> the yeast during the fermentations. Sometimes sugar levels are also<br />
considered. Although it is sometimes measured, the yield of extract is largely ignored.<br />
This is because most of the extracted materials in sorghum beers are derived from the<br />
adjunct(s) used. Malting losses, with seedling roots <strong>and</strong> shoots retained with the malt, are<br />
10±20% dry basis (Daiber <strong>and</strong> Taylor, 1995), but using some conditions they are very<br />
much higher (particularly if the seedling tissues are discarded, as is the case for malts<br />
intended for making lager beers), <strong>and</strong> so the maltster must strike a balance between malt<br />
quality <strong>and</strong> losses.<br />
The criteria that must be met are the level of diastatic power (DP), the free amino<br />
nitrogen (FAN), <strong>and</strong> a low tannin level (Daiber <strong>and</strong> Taylor, 1995; Daiber et al., 1973). In<br />
addition the beer should not have an appreciable level of mycotoxins or cyanide from the<br />
malt. Sorghum seedlings generate the cyanogenic glycoside dhurrin, which is<br />
enzymically degraded to glucose, p-hydroxybenzaldehyde <strong>and</strong> hydrogen cyanide (prussic<br />
acid). Toxic levels of prussic acid precursor occur in sorghum seedlings <strong>and</strong> some baking<br />
<strong>and</strong> steaming processes reduce this to low levels (Dada <strong>and</strong> Dendy, 1987). When<br />
sorghum was germinated for six days at three different temperatures the cyanide contents<br />
of the shoots increased continually to six days, to 614 ppm, at 25 ëC (77 ëF), peaked after<br />
two days germination at 666 ppm, at 30 ëC (86 ëF), <strong>and</strong> fell after two days at 35 ëC (95 ëF),<br />
when it was 385 ppm (Panasiuk <strong>and</strong> Bills, 1984). However, for reasons that are not clear,<br />
toxic levels of prussic acid do not appear in sorghum beers (Glennie, 1983).<br />
The grain chosen for malting must be of an acceptable variety, be clean <strong>and</strong><br />
undamaged <strong>and</strong> germinate well, usually at least 92±95%. As sorghum germinates the DP<br />
rises from a negligible value as both - <strong>and</strong> -amylases are synthesized in the embryo.<br />
There are variations, but generally 60±70% of the DP is due to the former enzyme. This<br />
contrasts with the situation found in barley malts, in which the DP is higher <strong>and</strong> the -<br />
amylase activity is relatively much higher (50±80% of the DP), <strong>and</strong> it exists preformed in<br />
the starchy endosperm (Briggs, 1998). There are problems in extracting the enzymes<br />
from some sorghum malts, due to the presence of tannins, <strong>and</strong> often 2% peptone is<br />
included in the extraction medium when DP is to be determined. -Glucosidase (maltase)<br />
is insoluble but active. Its importance in brewing is unclear <strong>and</strong> its activity is not<br />
determined on a routine basis. Its pH optimum is unusually acidic, pH 3.8 (Taylor <strong>and</strong><br />
Dewar, 1994) <strong>and</strong> so it is likely to act in mashing <strong>and</strong> produce the relatively high levels of<br />
glucose found in the worts.<br />
The DP levels of sorghum malts (20±60 SDU/g malt) are low compared to those of<br />
barley malts (150±200 SDU/g malt). The units of activity used, sorghum diastatic units<br />
(previously KDUs), are small <strong>and</strong> 1 SDU approximately equals 0.5 ëL. Many malts were<br />
<strong>and</strong> are produced with too low DP activities, often caused by growing the grain at too low<br />
temperatures. Commercial brewers now specify minimum DP values in malts, (often 28±<br />
33 SDU/g). Higher levels of enzymes <strong>and</strong> FAN in malts are obtained by malting highnitrogen<br />
grain <strong>and</strong> choosing smaller grains, which contain larger proportions of embryo<br />
tissues. In sorghum all the diastase is synthesized in the embryo, <strong>and</strong> apparently there is<br />
no contribution from the aleurone layer. Sorghum grains, unlike barley, wheat <strong>and</strong> some<br />
millets, do not respond to added gibberellic acid to any appreciable extent.<br />
The level of FAN in sorghum malt is important because this represents the nitrogencontaining<br />
nutrients needed by the microbes (Pickerell, 1986). The largest part of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
FAN is from the shoots <strong>and</strong> roots, which are ground up together with the rest of the grain,<br />
before mashing. FAN levels, like DP, increase with malting time. During malting<br />
proteolytic activity increases relatively little <strong>and</strong> this enzymic activity is poorly<br />
extractable but carboxypeptidase activity, which is concentrated in the embryo, increases<br />
substantially (Dewar et al., 1997c, d; Dewar <strong>and</strong> Taylor, 1995; Taylor, 1991). Both<br />
enzymic activities have acidic pH optima.<br />
The structure of the sorghum grain alters, i.e., undergoes modification, during malting<br />
but the cell walls of the endosperm appear to remain largely intact, while losing their<br />
physical strength. Modification can be assessed by measuring the porosity or<br />
compressibility of the grain (which both decline as modification proceeds) or, most<br />
simply, by measuring its specific gravity, which also declines (Daiber et al., 1973). These<br />
characteristics are not determined routinely.<br />
Relatively few studies have been reported aimed at optimizing sorghum malting for<br />
the production of African opaque beers. Other studies have been directed towards making<br />
sorghum malts for brewing clear, lager-type beers (Briggs, 1998). The variables that have<br />
been studied are steeping duration <strong>and</strong> temperature, the degree of sprinkling during<br />
germination, the temperature <strong>and</strong> duration of germination. The malt characteristics<br />
investigated have been the DP, FAN, hot water extract (total soluble solids) <strong>and</strong> malting<br />
loss. Extracts have been determined in various ways, e.g., after mashing for two hours at<br />
60 ëC (140 ëF). Values obtained after mashing at 45 ëC <strong>and</strong> then 70 ëC (113 <strong>and</strong> 158 ëF) are<br />
higher (Daiber <strong>and</strong> Taylor, 1995). In connection with brewing opaque beers malting<br />
losses are determined with the seedling tissues retained with the malt, so the losses<br />
encountered are mainly respiratory losses with a little due to leaching losses incurred<br />
during steeping.<br />
Early experiments indicated that malting temperatures of 20 ëC (68 ëF), or less, were<br />
much too low for maximum development of DP (Novellie, 1966). Systematic studies<br />
confirmed various other results, e.g. that for grain steeped <strong>and</strong> grown at about 30 ëC<br />
(86 ëF), 18 hours was the optimum steeping time <strong>and</strong> 4±5 days the best germination<br />
period for malting (Pathirana et al., 1983). In a study with germination times of up to six<br />
days, at temperatures of 24±36 ëC (75.2±96.8 ëF), with fixed initial steeping conditions<br />
but three levels of watering during germination, it was found that at steep out the<br />
moisture content was 33.7%, <strong>and</strong> after six days germination the moisture contents of the<br />
low, medium <strong>and</strong> highly wetted samples were 42.8%, 60.8% <strong>and</strong> 77% respectively<br />
(Morrall et al., 1986). Entire malt was used in this study <strong>and</strong> total soluble solids<br />
(`extract') were determined with two-hour mashes at 60 ëC (140 ëF), conditions which<br />
approximate to those used in mashing when making opaque beer. DP increased rapidly<br />
during the first four days of germination, the highest yield of 46.6 SDU/g being obtained<br />
after five days germination, at 24 ëC (75.2 ëF) in the medium-wetted grain, but 28 ëC<br />
(82.4 ëF) gave nearly as good results. At this stage the malting loss was 9.9%, the FAN<br />
was 129 mg/100 g malt <strong>and</strong> the total soluble solids value was 73.5%.<br />
At higher temperatures DP values were lower <strong>and</strong> under a number of conditions DP<br />
values peaked <strong>and</strong> then declined, as also happens in malting millets <strong>and</strong> wet-grown barley<br />
(Briggs, 1998). All the variables influenced FAN, which increased with increasing<br />
moisture content <strong>and</strong> germination time. The maximum level reached was 180 mg FAN/<br />
100 g malt, after six days germination, at the highest moisture content, in grain grown at<br />
32 ëC (89.6 ëF), when the malting loss was about 20%, the DP was 38 <strong>and</strong> the total soluble<br />
solids value was 66.1%. The maximum level of total soluble solids, 75.5%, was obtained<br />
from grain grown for six days at 24 ëC (75.2 ëF), with a medium moisture content, when<br />
the malting loss was 13.6%, the DP was 46.3 <strong>and</strong> the FAN was 148 mg/100 g malt.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Malting losses (seedling tissues retained with the malt) increased with germination time<br />
<strong>and</strong> increasing moisture content but were relatively little influenced by germination<br />
temperature.<br />
Clearly, strict control of malting conditions is necessary to produce malt in the best<br />
possible yield, with the desired specifications. Usually it will be necessary to compromise<br />
since, for example, conditions chosen to give maximum levels of FAN are associated<br />
with high malting losses <strong>and</strong> low DP values. Air-rests are beneficial when warm water<br />
steeping is tested (Ezeogu <strong>and</strong> Okolo, 1994, 1995). Further studies indicate that with<br />
steep aeration or air-rests longer steeping times, up to around 40 hours, are better than the<br />
usual shorter times found by earlier investigators (Dewar et al., 1997 a, b, c). Malt DP<br />
increased with steeping temperature up to 30 ëC (86 ëF), <strong>and</strong> of the temperatures tested,<br />
FAN <strong>and</strong> extract peaked at a steeping temperature of 25 ëC (77 ëF). Aeration increased the<br />
yields of extract <strong>and</strong> FAN. Using optimized steeping conditions, with air-rests, <strong>and</strong> two<br />
different wetting schedules, it was found that generally the optimum germination<br />
temperature was between 25 <strong>and</strong> 30 ëC (77 <strong>and</strong> 86 ëF). The roots <strong>and</strong> shoots contributed<br />
up to 61% of the whole malt FAN. In the winter, in South Africa, temperatures can fall<br />
below 18 ëC (64.4 ëF). It was concluded that much sorghum malt is made under seriously<br />
sub-optimal conditions.<br />
Ways have been sought for reducing the losses of dry matter that occur when sorghum<br />
is malted. Applications of potassium bromate <strong>and</strong> ammonia have not been beneficial, <strong>and</strong><br />
warm-water steeping, at 40 ëC (104 ëF), have given equivocal results, which varied with<br />
different grain samples (Ezeogu <strong>and</strong> Okolo, 1994, 1995). When dilute alkaline steeps<br />
(0.1% sodium hydroxide) were tested, probably with the initial object of extracting<br />
unwanted tannins, the moisture content <strong>and</strong> quality of the malts obtained were increased,<br />
even when low-tannin grains were treated (Dewar et al., 1997a, 1999; Ezeogo <strong>and</strong> Okolo,<br />
1999; Okolo <strong>and</strong> Ezeogu, 1996). However, there were varietal differences in responses to<br />
alkaline steeping.<br />
While various millets are malted in villages it seems unlikely that they are now malted<br />
commercially. The small sizes of the grains make them inconvenient to h<strong>and</strong>le <strong>and</strong>, in<br />
pneumatic malting plants, they tend to block the slots of the false floor <strong>and</strong> form dense<br />
layers that obstruct the passage of conditioning air. However, in the past millets were<br />
malted mixed with sorghum, which gave a more open bed of grain. It was believed that a<br />
more favourable mixture of enzymes was obtained from the mixture. Unlike sorghum,<br />
some millets respond to external doses of gibberellic acid. The studies carried out on the<br />
malting of millets have been in connection with the preparation of foodstuffs (Briggs,<br />
1998).<br />
16.3 <strong>Brewing</strong> African beers on an industrial scale<br />
The first attempts at brewing African beers on an industrial scale were made in southern<br />
Africa, using primitive equipment <strong>and</strong> without fully underst<strong>and</strong>ing the principles<br />
involved (e.g. Novellie, 1968; Oxford, 1926; Schwartz, 1956; Young, 1949). The failure<br />
rate was high. There have been great improvements in both underst<strong>and</strong>ing <strong>and</strong> brewery<br />
performance since the early years, <strong>and</strong> industrial scale brewing is spreading to other<br />
countries (Harris, 1997). Milled malts <strong>and</strong> adjuncts (maize grits, whole maize meal, degermed<br />
sorghum grits, or whole sorghum), may be delivered directly to the brewery. Of<br />
the several brewing systems in use the most common in South Africa is the Reef Process<br />
(Daiber <strong>and</strong> Taylor, 1995; Haggblade <strong>and</strong> Holzapfel, 1989; Harris, 1997; Novellie <strong>and</strong> De<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Schaepdrijver, 1986). In one version of this process milled sorghum malt (0.28 t;<br />
617.3 lb.) is slurried with water (25 hl; 550 imp. gal.) <strong>and</strong> is held at 48±50 ëC (118.4±<br />
122 ëF) for 8±18 hours. Older `sour', kept under conditions that maintain the lactic acid<br />
bacteria in a rapid, logarithmic state of growth, is mixed in to seed the mixture with<br />
thermophilic Lactobacilli. The process is stopped when the pH falls to 3.3 or less <strong>and</strong> the<br />
lactic acid concentration is about 0.8%.<br />
During this process some amylolysis <strong>and</strong> proteolysis occur, with increases in sugars<br />
<strong>and</strong> FAN. Alternatively, outside South Africa, commercially prepared lactic acid may be<br />
used to provide acidity. In the next stage, cooking, the sour is mixed with maize grits or a<br />
sorghum adjunct (2.3 t; 5,071 lb.) <strong>and</strong> water (168 hl; 3,696 imp. gal.) <strong>and</strong> the mixture, at<br />
pH 3.6±4.0, (depending on the product), is boiled for two hours at atmospheric pressure<br />
or for shorter times under pressure at temperatures up to 110 ëC (230 ëF). During the<br />
subsequent cooling a little malt may be added, at a temperature of about 80 ëC (176 ëF), to<br />
thin the material, by beginning to liquefy the starch so that transfer is easier. Water (13 hl;<br />
286 imp. gal.) <strong>and</strong> ground malt (0.62 t; 1,367 lb.) is mixed in <strong>and</strong> conversion takes place<br />
in this cooled, acidic mash, which is continued at 60 ëC (140 ëF) for two hours, at a pH of<br />
3.6±3.8. Under these conditions the granular maize <strong>and</strong> sorghum starch does not<br />
gelatinize <strong>and</strong> the activities of the amylases are limited by the low pH so sugar production<br />
is limited. Proteolysis occurs <strong>and</strong> the level of FAN increases to a useful extent. The<br />
thinned mash is then `strained', at 60 ëC (140 ëF), either through a screen, or by<br />
centrifugation followed by passage through a vibrating screen. The collected strainings<br />
weigh about 3 t, 6,614 lb., <strong>and</strong> contain about 1.008 t, 2,222 lb., of dry material, which<br />
represents about 30% of the grist solids. The screenings are sold, wet or after drying, for<br />
cattle food. This wasteful process has proved difficult to improve. Although it is possible<br />
to re-mash the screenings with added enzymes to obtain a secondary wort, this finding<br />
seems not to have been exploited. The objective of straining is to remove coarse particles<br />
of more than about 0.25 mm (about 0.01 in.) width.<br />
After straining the `wort', about 200 hl, 4,399 imp. gal., cooled to 28±30 ëC (82.4±<br />
86 ëF), is transferred to a stainless steel fermenter of 150 or 270 hl (3,300 or 5,939 imp.<br />
gal.) capacity, <strong>and</strong> is pitched with a selected, pure, dried culture yeast (5.5 kg; 12.1 lb.).<br />
At this stage the wort contains 6±7% fermentable sugars (glucose, maltose <strong>and</strong><br />
maltotriose in the approximate ratio 1:3:1), 3% soluble dextrins, more than 1% of<br />
gelatinized starch <strong>and</strong> about 2% ungelatinized starch. The optimal conditions for<br />
proteolysis are 51 ëC (123.8 ëF) <strong>and</strong> pH 4.6, <strong>and</strong> so it is not surprising that FAN increases<br />
during mashing. At the end of the mash about 30% of the FAN was generated during<br />
mashing, the other 70% being from the malt <strong>and</strong> having been generated during souring.<br />
As malts are prepared with higher DP values the ratio of adjunct to malt tends to increase,<br />
carrying with it the risk that the mash will contain too little FAN to adequately support<br />
the yeast growth during fermentation. Perhaps 100 mg FAN/litre is the minimum safe<br />
concentration, when the FAN is derived from sorghum malt. Fermentation proceeds for a<br />
selected time (usually 8±24 h), then the actively fermenting beer is sold either on draught<br />
or in small waxed cardboard or larger polyethylene containers. All containers are vented<br />
to allow the escape of the continuously generated carbon dioxide. Unfortunately this is<br />
apt to allow beer to escape as well.<br />
Compared to `clear beers' opaque beers are viscous (e.g. 15 mPa.s) <strong>and</strong> are rich in<br />
fusel oils. Other analyses of commercial South African beers (ranges, with the mean<br />
values in brackets) are: alcohol, 2.4±4.0% w/w (3%), total solids 2.6±7.2%,w/w (4.9);<br />
insoluble solids, 1.6±4.3% (2.3); lactic acid, 164±250 mg% (213); pH3.2±3.9 (3.5);<br />
volatile acids as acetic acid g/100ml, 0.012±0.029 (0.026), total nitrogen 0.065±0.115%<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(0.084) (Novellie, 1968; Novellie <strong>and</strong> De Schaepdrijver, 1986). Other reported values<br />
differ significantly. Part of the art in this brewing system is to produce the desired levels<br />
of nutrients for the lactic acid bacteria <strong>and</strong> the yeast, to gelatinize some of the starch, but<br />
not to degrade the starch so much that the beer is `thin'. The residual starch is relatively<br />
poor in amylose <strong>and</strong> the side chains of the amylopectin are roughly halved in length<br />
(Glennie, 1988).<br />
The iJuba process gives a less viscous beer, which is made to meet the preferences<br />
of the Zulu people. Starch conversion is carried further in this process. Water (165 hl;<br />
3,630 imp. gal.), maize grits (1.275 t; 2,811 lb.), ground sorghum grain (0.795 t;<br />
1,753 lb.) <strong>and</strong> sorghum malt (`pre-malt'; 0.105 t; 232 lb.), or a microbial amylase, are<br />
combined <strong>and</strong> the mixture is heated <strong>and</strong> finally boiled, for two hours, at its natural pH,<br />
when enzyme activity ceases, starch is gelatinized <strong>and</strong> microbes are almost all<br />
destroyed. The pre-malt or microbial amylase is added to liquefy some of the starch<br />
<strong>and</strong> reduce subsequent h<strong>and</strong>ling problems. The mix is cooled, to 60 ëC (140 ëF), <strong>and</strong><br />
sorghum malt (1.1 t; 2,425 lb.) is added <strong>and</strong>, after two hours conversion, which occurs<br />
more rapidly than in the Reef process, at this `natural' pH, water is added <strong>and</strong> the mash<br />
is allowed to sour for around four hours at 50 ëC (122 ëF) to a pH of 3.8±4. The souring<br />
may be initiated by the addition of a pure culture of Lactobacillus delbruÈckii<br />
(leichmannii) or part of a previous sour. A small amount of ground malt (0.105 t;<br />
232 lb) is then added <strong>and</strong> the mixture is heated to boiling, when the lactic acid bacteria<br />
are killed <strong>and</strong> all enzyme activity is terminated. Following this heat treatment the<br />
mixture is cooled to 40 ëC (104 ëF), more sorghum malt (0.105 t; 232 lb.) is added, <strong>and</strong><br />
the mixture is strained <strong>and</strong> cooled to 28 ëC (82.4 ëF). Strainings amount to 2.5 t, 5,512<br />
lb., less than in the Reef process because starch liquefaction is more advanced <strong>and</strong> so<br />
the viscosity of the mixture is less, allowing a `cleaner' separation of the coarser<br />
materials. Active dried yeast (3.5 kg; 7.72 lb.) is added to the wort (about 200 hl; 4,400<br />
imp. gal.) <strong>and</strong> fermentation proceeds for 8±48 h, at 28 ëC (82.4 ëF), before the<br />
fermenting beer is packaged <strong>and</strong> distributed. In contrast to the Reef process the<br />
amylases convert the mash at the natural pH, when they are much more active than in<br />
the acid conditions used in the Reef process mash. In addition the heat treatment<br />
effectively sterilizes the mash before the last malt addition, straining <strong>and</strong> pitching.<br />
The Kimberley style of brewing is intermediate between the Reef <strong>and</strong> the iJuba styles,<br />
<strong>and</strong> was developed to allow brewing with poor-quality sorghum malts with low DP<br />
values. The first stages involve two process streams. In the `sour stream' sorghum malt<br />
(0.9 t; 1,984 lb.) is mixed with water (25 hl; 550 imp. gal.) <strong>and</strong> the mixture is held at about<br />
49 ëC (120.2 ëF) for 18 h, when the pH falls to 3.2. The microbes are then killed, <strong>and</strong><br />
enzymes are inactivated, by heating the sour to 85 ëC (185 ëF). In the `main' process<br />
stream maize grits (2.3 t; 5,071 lb.) <strong>and</strong> water (168 hl; 3,696 imp. gal.) are mixed <strong>and</strong><br />
sometimes some microbial -amylase is added to `thin' the starch as it gelatinizes. The<br />
mixture is then boiled for two hours, at its natural pH. The cooked material is mixed with<br />
water (13 hl; 286 imp. gal.) <strong>and</strong> ground sorghum malt (0.62 t; 1,367 lb.), <strong>and</strong> conversion<br />
proceeds for 2 h at 60 ëC (140 ëF) at the `natural' pH, 5.8±6.0. After mixing in the heated<br />
sour from the other process stream, when the fall in pH checks the activities of the<br />
amylases, the mixture is strained, yielding 3 t, 6,614 lb., of strainings, <strong>and</strong> the wort is<br />
pitched with 5.5 kg, 12.1 lb., of dried yeast. Fermentation <strong>and</strong> distribution are carried out<br />
in the usual ways.<br />
While the three processes described, including minor variations of them, are the most<br />
usual others are also used (Diefenbach, 1996; Harris, 1997). In some of these processes<br />
`sours' are replaced with lactic acid. In the split sour Chibuku process maize <strong>and</strong> sorghum<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
grits are cooked together with some of a sorghum malt sour (about 40%), <strong>and</strong> more sour<br />
(60%) is added after a conversion stage, which is carried out on the cooked material after<br />
the addition of sorghum malt <strong>and</strong> more water. After heating to 80 ëC (176 ëF), at pH 3.7,<br />
the material is subjected to a second conversion stage, with added fungal amyloglucosidase,<br />
for 30 minutes at 60 ëC (140 ëF). The amyloglucosidase degrades starch <strong>and</strong><br />
dextrins to glucose. The next stages are straining, cooling <strong>and</strong> fermentation.<br />
A predictable problem is continuing activity of the amyloglucosidase in the beer. In<br />
the Chibuku Zimbabwe/Botswana process -amylase <strong>and</strong> lactic acid are added to the<br />
maize <strong>and</strong> water cook <strong>and</strong> amyloglucosidase <strong>and</strong> sorghum malt are used in a second<br />
conversion stage, the enzymes needed in the first stage being provided by sorghum malt.<br />
So in this process milled maize (1.8 t; 3,968 lb.) is mixed with water (95 hl; 2,090 imp.<br />
gal.), lactic acid (8 litres; 1.76 imp. gal., 80%), <strong>and</strong> amylase. After a boil for 1.5 h at<br />
pH 5.0, water (60 hl; 1,320 imp. gal.) is added <strong>and</strong> sorghum malt (0.46 t; 1,014 lb.) is<br />
added to the cooled mixture <strong>and</strong> conversion takes place for two hours, at 60 ëC (140 ëF)<br />
<strong>and</strong> pH 5.5. After centrifugal straining the wort is pasteurized by heating to 80 ëC<br />
(176 ëF). The mixture is cooled to 60 ëC (140 ëF) <strong>and</strong> is incubated with amyloglucosidase<br />
<strong>and</strong> more sorghum malt (0.02 t; 44.1 lb.) for 0.5 h. Following this second conversion the<br />
wort is cooled <strong>and</strong> pitching <strong>and</strong> fermentation are carried out, at 26 ëC (78.8 ëF). In the<br />
Chibuku/Zimbabwe/Malawi process malt is not used at all. Extract is derived exclusively<br />
from milled maize. Acidity is provided by lactic acid <strong>and</strong> the enzymes employed are<br />
microbial -amylase <strong>and</strong> amyloglucosidase. The product, by usual criteria, is not a beer<br />
but a beer substitute (Harris, 1997).<br />
Comparisons between `home' <strong>and</strong> industrial opaque beer brewing <strong>practice</strong>s <strong>and</strong><br />
between these <strong>and</strong> the methods used in making `clear beers' are interesting, but opaque<br />
beer home brewing does not seem to have been studied scientifically. Obvious<br />
differences are the scales of operation <strong>and</strong> the near total lack of control, other than the<br />
brewster's judgement, in home brewing. Industrial brewers have better raw materials of<br />
more uniform quality so, for example, they specify that malts should have DPs of 28±35<br />
SDUs <strong>and</strong> water contents of less that 10%. The more exact temperature control available<br />
in industrial breweries has advantages at every stage of operation. The souring process is<br />
under better control, both because by operating at around 50 ëC (122 ëF) only<br />
thermophilic bacteria are encouraged to grow, <strong>and</strong> inoculation may be with a pure<br />
culture of a Lactobacillus, <strong>and</strong> because pH <strong>and</strong> acidity can be measured. Not all<br />
thermophilic bacteria are desirable in the sour. In home brewing the souring occurs in a<br />
cooling environment, so many different bacteria may grow, thermophiles being<br />
succeeded by mesophiles, with the generation of less lactic acid <strong>and</strong> the formation of<br />
unwanted flavours <strong>and</strong> aromas.<br />
In home brewing the souring <strong>and</strong> alcoholic fermentations are not clearly separated,<br />
rather they overlap. The lactic acid is said to help the softening of the endosperm during<br />
cooking, so facilitating the release <strong>and</strong> gelatinization of the starch granules. Final beers<br />
should contain both gelatinized starch <strong>and</strong> ungelatinized granules. Starch gelatinization<br />
temperature ranges are maize, 62±74 ëC (143.5±165.2 ëF), sorghum, 69±75 ëC (156.2±<br />
167 ëF), millets, 54±85 ëC (129.2±176 ëF) <strong>and</strong> barley, 60±62 ëC (140±143.6 ëF) (Briggs,<br />
1998). Consequently in mashes carried out at 60 ëC (140 ëF), the sorghum amylases will<br />
not attack granular sorghum, millet or maize starches to any appreciable extent. In several<br />
of the mashing processes the acidic pH values (pH 3.6±4.0) limit the activities of the<br />
sorghum amylases ( -amylase optimum pH 4.5±5; -amylase optimum 5.2±5.5; Daiber<br />
<strong>and</strong> Taylor, 1995), acting on the gelatinized starch. As with all -amylases, calcium ions<br />
form an integral part of the sorghum enzyme <strong>and</strong> inclusion of calcium salts in the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
mashing liquor stabilizes the enzyme <strong>and</strong> enhances its activity, as is the case with the<br />
barley enzyme during mashes for making clear beers (Taylor, 1989, 1992).<br />
Commercially, the alcoholic fermentation is initiated by pitching with pure, top<br />
fermenting yeast, rather than the mixture of microbes present on the sorghum malt <strong>and</strong><br />
brewing vessels used in home brewing, which give unpredictable fermentations. Even so,<br />
industrial fermentations contain microbes from the raw materials <strong>and</strong> these are major<br />
contributors to spoilage. During the alcoholic fermentation acidification continues <strong>and</strong> so<br />
the pH continues to fall. Hetero-fermentative, mesophilic bacteria produce some lactic<br />
acid <strong>and</strong> also a range of other products, with adverse effects on flavour. As fermentation<br />
slackens, <strong>and</strong> the generation of carbon dioxide declines, air gains more ready access to<br />
the liquid, oxidative changes can occur <strong>and</strong> some alcohol is oxidized to acetic acid, which<br />
confers an unwanted, vinegary flavour <strong>and</strong> aroma. Yeast may die <strong>and</strong> autolyse, reducing<br />
the competition with, <strong>and</strong> supplying nutrients for, contaminating microbes. Some<br />
microbes may ultimately form a pellicle on the surface of the beer. Thus spoilage is<br />
inevitable in several days <strong>and</strong> this creates major problems for supplying beers over long<br />
distances <strong>and</strong> in meeting the variable dem<strong>and</strong>s for beer, which peak at weekends <strong>and</strong> at<br />
the ends of each calendar month.<br />
16.4 Attempts to obtain stable African beers<br />
The irregular <strong>and</strong> short shelf-lives of African beers, say three days in summer <strong>and</strong> five<br />
days in winter in South Africa, create problems of supply <strong>and</strong> distribution <strong>and</strong> ensure that<br />
breweries have excess brewing capacity for much of the time. This they must have to<br />
allow them to meet the high dem<strong>and</strong>s at weekends. In an older type of brewery the shelflife<br />
of the product may be three days, while in a modern, scrupulously cleaned, dust<br />
controlled plant, using pure cultures of yeast <strong>and</strong> perhaps of Lactobacilli, the shelf-life<br />
may still be only five days. Various attempts have been made to produce stable beers but<br />
with only limited success. Experimentally, a stable form of Kenyan beer was prepared by<br />
pasteurization <strong>and</strong> a clear Nigerian beer, oyokpo, was stabilized with benzoic acid<br />
(Daiber <strong>and</strong> Taylor, 1995; Iwuagwu <strong>and</strong> Izuagbe, 1985; Nout, 1980). Most studies<br />
concern southern Africa (Harris, 1997; Haggblade <strong>and</strong> Holzapfel, 1989; Novellie <strong>and</strong> De<br />
Schaepdrijver, 1986). Chilling the wort to 14±16 ëC (57.2±60.8 ëF) before pitching with<br />
yeast slows the fermentation, <strong>and</strong> lengthens the shelf-life of the beer. However, this<br />
approach has problems as the product is consumed warm by choice <strong>and</strong> refrigerated<br />
transport <strong>and</strong> dispensing equipment is not available. Storing under a top pressure of<br />
carbon dioxide has been proposed, but apparently this is not used. Heavily -irradiating<br />
the malt greatly reduces the population of microbes <strong>and</strong> increases the shelf-life of beer<br />
from about four to six days. Again, this process seems not to be used.<br />
Successful ways around the instability problem involve the use of beer powders or<br />
concentrated worts. Techniques using pasteurization of worts or beers have not worked so<br />
well. Beer powders consist of dry, finely milled sorghum malt, dry yeast <strong>and</strong> pre-cooked<br />
maize meal. The cooking is by steam injection, which gelatinizes at least some of the<br />
starch. Some formulations may also contain some acidified material. `<strong>Brewing</strong>' consists<br />
of dispersing the powder in warm water. Conversion, acidification (by bacteria from the<br />
malt) <strong>and</strong> alcoholic fermentation begin rapidly <strong>and</strong> simultaneously. The product is ready<br />
for consumption in 4±8 h <strong>and</strong> has a shelf-life of about one day. This process resembles the<br />
production of the millet beer, busaa (Nout, 1980). The beer is of poor quality. In part this<br />
is because initially the acidity is low, giving spoilage organisms a chance to multiply<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
efore the pH falls. The powder is stable <strong>and</strong> so can easily be transported <strong>and</strong> stored,<br />
allowing beer to be produced in remote locations, at short notice <strong>and</strong> in small or large<br />
amounts. Another approach is to make wort, with a low solids content, in an approved<br />
fashion <strong>and</strong> then, after straining, concentrating it in a film evaporator to around 50%<br />
solids (Harris et al., 1999). It is possible to spray-dry this material. The concentrated wort<br />
is syrupy <strong>and</strong> contains solids. It is relatively stable, even to spoilage by osmophilic yeasts.<br />
Because it is concentrated <strong>and</strong> stable it can be widely distributed relatively easily. To<br />
brew syrup, 5 kg (11 lb.), is diluted to 25 litres (5.5 imp. gal.) <strong>and</strong> is pitched with brewing<br />
yeast. The product can be a high-quality beer.<br />
Both batch- <strong>and</strong> tunnel-pasteurization treatments failed, partly because the heating<br />
gelatinized the granular starch in the beer, making it excessively viscous. Partial<br />
microbiological control was obtained by flash-pasteurizing wort at 72±76 ëC (161.6±<br />
168.8 ëF) for 12±15 s using plate heat exchangers. Viscosity <strong>and</strong> flavour were scarcely<br />
altered, but because of the solids present complete sterilization was not achieved. Beer<br />
could be flash-pasteurized more successfully, at 75±80 ëC (167±176 ëF) for 20±25 s.<br />
However, all the carbon dioxide was removed <strong>and</strong> so the beer was `flat', that is, it lacked<br />
effervescence <strong>and</strong> `tingle'. Carbonating this product failed, apparently because of<br />
difficulties caused by the high solids content. This flat beer has limited acceptability but<br />
some is sterile packaged <strong>and</strong> sold. Maintaining sterility has been difficult but where it has<br />
held the product has a long shelf-life. Another approach has been to bottle pasteurized<br />
beer with added yeast <strong>and</strong> a calculated amount of sugar, (a process closely similar to the<br />
traditional British process of conditioning `in bottle' with added priming sugars <strong>and</strong> a<br />
secondary yeast). In the South African process `in bottle conditioning' occurs as the yeast<br />
ferments for 5±12 days <strong>and</strong> the pressure of carbon dioxide rises to 20±30 psi, being<br />
limited by the amount of sugar added. The beer has an unusual flavour but it is acceptable<br />
<strong>and</strong> relatively stable (Harris, 1997). Bottles are said to have burst, possibly because of<br />
over-dosing with sugar, permitting an excessive accumulation of carbon dioxide, with a<br />
consequent excessive rise in pressure.<br />
16.5 Beer composition <strong>and</strong> its nutritional value<br />
Opaque beers are valuable foodstuffs, <strong>and</strong> are the best of all the alcoholic beverages in<br />
this respect. However, there is no reliable way to decide exactly how important they are<br />
in <strong>practice</strong>. There are several reasons for this. Home-brewed beers almost certainly vary<br />
very greatly in their compositions (but these are unknown) <strong>and</strong> the natures <strong>and</strong><br />
proportions of the raw materials used, the quantities of beer drunk, when (consumption<br />
varies with the season) <strong>and</strong> by whom, <strong>and</strong> the nutritional status of the drinkers all vary<br />
(Daiber <strong>and</strong> Taylor, 1995; Haggblade <strong>and</strong> Holzapfel, 1989; Heerden, van, 1989a, b;<br />
Novellie <strong>and</strong> De Schaepdrijver, 1986). Many years ago it was noted that in some seasons<br />
the men of particular tribes lived on beer alone for extended periods. As there is a<br />
gradation between soured, acidified, essentially alcohol-free porridges <strong>and</strong> acidified<br />
beers, <strong>and</strong> as the alcohol contents of the beers are variable <strong>and</strong> are continually increasing<br />
as fermentation continues, no conclusions regarding alcohol consumption can be drawn.<br />
With commercial beers the alcohol contents, at the time of sale, are relatively low<br />
(generally 3% or less) compared to clear beers (3±9%) <strong>and</strong> very much less than in wines<br />
(11±13%) or spirits (often 40%).<br />
The nutrients present in beers, particularly the vitamins, are derived from the raw<br />
materials, Lactobacilli <strong>and</strong> the yeast. Some of the materials initially present in the wort<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
are taken up by the Lactobacilli <strong>and</strong> the yeast <strong>and</strong> may be converted into different<br />
substances. Some nutrients are destroyed by the brewing process, while those that are<br />
present in the beer may not be `bio-available' <strong>and</strong> cannot be used by the consumer. The<br />
compositions of several commercial beers have been reported. Ranges (<strong>and</strong> mean values)<br />
for eight beers were: energy content, in kJ/litre, 1,530±1,840 (1,651); alcohol content, in<br />
g/litre, 16±32 (25.4); crude protein, g/litre, 4±9 (5.4); fat, g/litre, trace 1 (trace); crude<br />
fibre, g/litre, trace 1 (trace); ash, g/litre, 1±2 (1.13); carbohydrate, g/litre, 32±59 (47.6).<br />
Starch contents of five beers ranged between 27.8 to 32.7 g/litre (mean 29.7) <strong>and</strong> the<br />
values for three vitamins in 15 beers, all in mg/litre, were thiamine 0.13±0.36 (0.24);<br />
riboflavin, 0.30±0.47 (0.39) <strong>and</strong> for nicotinic acid, 2.32±3.74 (2.93) (Heerden, van,<br />
1989b). Some contents of mineral elements, ranges <strong>and</strong> mean values all in mg/litre were<br />
potassium, 145±438 (280); sodium, 7±46 (21); calcium, 22±57 (38); magnesium, 69±174<br />
(111); copper, 0.11±0.26 (0.18); iron, 0.9±2.1 (1.4); manganese, 0.8±2.2 (1.4) <strong>and</strong> zinc,<br />
1.0±1.9 (1.4) (Novellie <strong>and</strong> De Schaepdrijver, 1986). Phosphorus contents, in other<br />
estimates on five beers, in mg/litre, were 96±565 (218).<br />
Although particular substances are present in beers it does not follow that they are<br />
available to the imbiber. The living yeast in opaque beers takes up many vitamins quickly<br />
<strong>and</strong> these, together with those already present in the micro-organisms, are then largely<br />
unavailable. Heat treating or pasteurizing the beer damages or kills the yeast <strong>and</strong> the<br />
vitamins are released into solution <strong>and</strong> are then available. Analysis of beer crude protein<br />
shows that a significant amount is present, with a `plant-like' distribution of amino acids,<br />
which do not contain an ideal mixture of nutritionally essential amino acids but which,<br />
nevertheless, has a good proportion of the essential amino acid lysine, which is probably<br />
derived from the yeast. As the lysine may be largely contained within the yeast cells, its<br />
availability is in doubt.<br />
Some phosphorus could be present as phytic acid (meso-inositol hexaphosphate,<br />
(4.156)), a substance with strong chelating properties that can limit the nutritional<br />
availability of metal ions. However, <strong>and</strong> in contrast to many other cereal-based products,<br />
phytate was not detected in significant amounts in these beers. The replacement of<br />
sorghum grain adjunct with refined maize grits (from which the nutrient-rich embryos<br />
<strong>and</strong> aleurone layers have been removed) in both commercial <strong>and</strong> home brewing, <strong>and</strong> the<br />
progressive reduction in the proportion of sorghum malt used in the grist in commercial<br />
brewing (facilitated by the increasing DP values of sorghum malts <strong>and</strong> the<br />
supplementation of these malts with microbial amylases) are changes which have<br />
reduced the nutritional values of beers, in particular the B vitamin contents, <strong>and</strong> have led<br />
to the proposal that maize grits should be replaced with the traditional whole sorghum<br />
grain <strong>and</strong> the beers be supplemented with vitamins, thiamine <strong>and</strong> vitamin C (ascorbic<br />
acid; Heerden, van, 1989 a, b). Home-brewed beers, made with higher proportions of<br />
sorghum malt than are used in commercial brewing, are probably nutritionally superior.<br />
In addition to the B vitamin content these beers almost certainly contain higher levels of<br />
desirable crude fibre since the `home' straining procedures are less stringent than those<br />
used in commercial brewing. Probably each litre of commercial beer consumed can<br />
provide 10% of the daily recommended protein requirement <strong>and</strong> 14% of the energy<br />
requirement of a moderately active man <strong>and</strong> is a useful source of the B vitamins thiamine,<br />
riboflavin <strong>and</strong> nicotinic acid as well as the minerals iron, zinc, manganese, magnesium<br />
<strong>and</strong> phosphorus. The starch is a good energy source but the alcohol is less good. Drinkers<br />
usually consume about two litres (3.52 imp. pints) of opaque beer/day.<br />
Specifications for commercial beers are likely to include values for total solids, crude<br />
protein <strong>and</strong> lactic acid contents, pH, alcohol content when sold, (3% or less in South<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Africa), <strong>and</strong> an upper limit on volatile acids (as acetic acid). In addition the products must<br />
be free of pathogenic organisms, must not have a tendency for the solids to separate <strong>and</strong><br />
precipitate, have an appropriate viscosity, appearance, (colour, opacity <strong>and</strong> foam), <strong>and</strong> be<br />
acceptable as judged by a taste panel.<br />
16.6 References<br />
BRIGGS, D. E. (1998) Malts <strong>and</strong> Malting. Aspen Publishing, Inc., Gaithersburg, MD, 796 pp.<br />
CHEVASSUS-AGNES, S., FAVIER, J. C. <strong>and</strong> ORSTOM, A. J. (1976) Cahiers Nutr. Diet., 11 (2), 89.<br />
DADA, L. O. <strong>and</strong> DENDY, D. A. V. (1987) Tropical Sci., 27, 101.<br />
DAIBER, K. H. (1975) J. Sci. Fd. Agric., 26, 1399.<br />
DAIBER, K. H. (1978) Special Report BB114. Manual on the treatment <strong>and</strong> malting of birdproof grain<br />
sorghum. The Sorghum Beer Unit, CSIR, South Africa. 12 pp.<br />
DAIBER, K. H. <strong>and</strong> TAYLOR, J. R. N. (1995) in Sorghum <strong>and</strong> Millets: Chemistry <strong>and</strong> Technology. (Dendy,<br />
D. A. V. ed.), A. A. C. C. St Paul, Minn. p. 299.<br />
DAIBER, K. H., MALHERBE, L. <strong>and</strong> NOVELLIE, L. (1973) Brauwissenschaft, 26 (7), 220, 248.<br />
DENDY, D. A. V. (ed.) (1995) Sorghum <strong>and</strong> Millets; chemistry <strong>and</strong> Technology. Amer. Assoc. Cereal<br />
Chemists. St. Paul, Minn.<br />
DEWAR, J. <strong>and</strong> TAYLOR, J. R. N. (1995) Proc. 5th Sci. Tech. Conv. Inst. of <strong>Brewing</strong>, (Central <strong>and</strong> Southern<br />
African Sect.), Victoria Falls, p. 93.<br />
DEWAR, J., OROVAN, E AND TAYLOR, J. R. N. (1997a) J. Inst. <strong>Brewing</strong>, 103, 283.<br />
DEWAR, J., TAYLOR, J. R. N. <strong>and</strong> BERJAK, P. (1997b) Proc. 6th Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong><br />
Southern African Sect.), Durban, p. 29.<br />
DEWAR, J., TAYLOR, J. R. N. <strong>and</strong> BERJAK, P. (1997c) J. Cereal Sci., 26, 129.<br />
DEWAR, J., TAYLOR, J. R. N. <strong>and</strong> BERJAK, P. (1997d) J. Inst. <strong>Brewing</strong>, 103, 171.<br />
DEWAR, J., DONALDSON, S. <strong>and</strong> TAYLOR, J. R. N. (1999) Proc. 7th Conv. Inst of <strong>Brewing</strong> (African Sect.),<br />
Nairobi, p. 217.<br />
DIEFENBACH, M. (1996) Brauwelt Internat., 14 (5), 431.<br />
DIRAR, H. A. (1978) J. Food Sci., 43, 1983.<br />
EKUNDAYO, J. A. (1969) J. Food Sci., 4, 217.<br />
EZEOGU, L. I. <strong>and</strong> OKOLO, B. N. (1994) J. Inst. <strong>Brewing</strong>, 100, 335.<br />
EZEOGU, L. I. <strong>and</strong> OKOLO, B.N. (1995) J. Inst. <strong>Brewing</strong>, 101, 39.<br />
EZEOGU, L. I. <strong>and</strong> OKOLO, B. N. (1999) J. Inst. <strong>Brewing</strong>, 105, 49.<br />
FAPARUSI, S. I. (1970) J. Sci. Food Agric., 21, 79.<br />
FAPARUSI, S. I., OLOFINBOBA, M. O. <strong>and</strong> EKUNDAYO, J. A. (1973) Zeits. f. Allg. Mikrobiologie, 13 (7), 563.<br />
FOX, F. W. (1938) J. S. African Chem. Inst., 21, 39.<br />
GLENNIE, C. W. (1983) J. Agric. Food Chem., 31, 1295.<br />
GLENNIE, C. W. (1988) Starch/StaÈrke, 40 (7), 259.<br />
HAGGBLADE, S. <strong>and</strong> HOLZAPFEL, H. (1989) in Industrialization of Indigenous Fermented Foods<br />
(Steinkraus, K. H. ed.). Marcel Dekker Inc. New York, p. 191.<br />
HARRIS, R. N. (1997) Proc. 6th Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.), Durban,<br />
p. 89.<br />
HARRIS, R. N., MATHIBA, K., JOUSTRA, S. M., VILJOEN, C. R. <strong>and</strong> YENKETSWAMY, C. (1999) Proc. 7th Conv.<br />
Inst. of <strong>Brewing</strong> (African Sect.), Nairobi, p. 218.<br />
HEERDEN, I. V. VAN (1989a) J. Inst. <strong>Brewing</strong>, 95, 17.<br />
HEERDEN, I. V. VAN (1989b) Proc. 2nd Sci. Tech. Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern African<br />
Sect.), Johannesburg, p. 293.<br />
IWUAGWU, Y. O. U. <strong>and</strong> IZUAGBE, Y. S. (1985) J. Appl. Bacteriol., 59, 487.<br />
MIRACLE, M. P. (1965) in Food Technology the World Over (Peterson, M. S. ed.). Avi Publishing,<br />
Westport, Conn. p. 107.<br />
MORCOS, S. R., HEGAZI, S. M. <strong>and</strong> EL-DAMHOUGY, S. T. (1973) J. Sci. Food Agric., 24, 1157.<br />
MORRALL, P., BOYD, H. K., TAYLOR, J. R. N. <strong>and</strong> VAN DER WALT, W. H. (1986) J. Inst. <strong>Brewing</strong>, 92, 439.<br />
NOUT, M. J. R. (1980) Chem. Mikrobiol. Technol. Lebensm., 6, 137, 174.<br />
NOUT, M. J. R. <strong>and</strong> DAVIES, B. J. (1982) J. Inst. <strong>Brewing</strong>, 88, 157.<br />
NOVELLIE, L. (1966) Internat. Brew. Distill., 1 (1), 27.<br />
NOVELLIE, L. (1968) Wallerstein Labs. Communs., 31, 17.<br />
NOVELLIE, L. (1977) in Proc. of a Symp.on Sorghum <strong>and</strong> Millets for Human Foods. (Dendy, D. A. V. ed.),<br />
The Tropical Products Institute, London. p. 73.<br />
NOVELLIE, L. <strong>and</strong> DE SCHAEPDRIJVER, P. (1986) in Progress in Industrial Microbiology. 23. Microorganisms<br />
in the production of food. (Adams, M. R. ed.), Elsevier, Amsterdam, p. 73.<br />
NZALIBE, H. C. <strong>and</strong> NWASIKE, C. C. (1995) J. Inst. <strong>Brewing</strong>, 101, 345.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
OGUNDIWIN, J. O. (1977) Brew. Distill. Internat., 7 (6). 40.<br />
OKOLO, B. N. <strong>and</strong> EZEOGU, L. I. (1996) J. Inst. <strong>Brewing</strong>, 102, 79, 277.<br />
O'ROURKE, T. (2001) Brewer Internat., 1 (10), 46.<br />
OXFORD, T. (1926) J. Inst. <strong>Brewing</strong>, 32, 314.<br />
PANASIUK, O. <strong>and</strong> BILLS, D. D. (1984) J. Food Sci., 49, 791.<br />
PATHIRANA, R. A., SIVAYOGASUNDARAM, K. <strong>and</strong> JAYATISSA, P. M. (1983) J. Food Sci. Technol., 20 (3),<br />
108.<br />
PERISSE, J., ADRIAN, J., REVET, A. <strong>and</strong> LE BARREÂ , S. (1959) Ann. de la Nutrit. et de l'Aliment., 13 (1), 1.<br />
PETERSON, M. S. <strong>and</strong> TRESSLER, D. H. (1965) Food Technology the World Over. 2. South America, Africa,<br />
Middle East <strong>and</strong> Asia, Avi Publishing, Westport, Conn. p. 130.<br />
PICKERELL, A. T. W. (1986) J. Inst. <strong>Brewing</strong>, 92, 568.<br />
SCHWARTZ, H. N. (1956) J. Sci. Food Agric., 7, 101.<br />
TAYLOR, J. R. N. (1989) Proc. 2nd Sci. Tech. Conv. Inst. of <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.),<br />
Johannesburg, p. 275.<br />
TAYLOR, J. R. N. (1991) Proc. 3rd Sci. Tech. Conv. Inst of <strong>Brewing</strong> (Central <strong>and</strong> Southern African Sect.),<br />
Victoria Falls, p. 18.<br />
TAYLOR, J. R. N. (1992) J. Amer. Soc. Brew. Chem., 50, 13.<br />
TAYLOR, J. R.N. <strong>and</strong> DEWAR, J. (1994) J. Inst. <strong>Brewing</strong>, 100, 417.<br />
YOUNG, R. S. (1949) J. Inst. <strong>Brewing</strong>, 55, 371.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
17<br />
Microbiology<br />
17.1 Introduction<br />
Thedesignofthebreweryplant<strong>and</strong>operationoftheprocessmustbesuchthatproduction<br />
yeast strains remain segregated with no possibility of inter-mixing. In addition,<br />
contamination with foreign micro-organisms must be prevented. These goals can be<br />
achieved by ensuring high st<strong>and</strong>ards of hygiene within the brewery. Brewery plant is<br />
constructedfromstainlesssteelforeaseofcleaning.Themodernprocesstendstobefully<br />
enclosed to ensure that amicrobiological barrier is maintained between process liquids<br />
<strong>and</strong> the external environment. All critical parts of the plant are fitted with automatic <strong>and</strong><br />
efficient cleaning in place (CIP) systems. The microbiological integrity of the brewing<br />
process must be confirmed with appropriate testing. For acomplex process such as<br />
brewing this necessitates the adoption of asampling plan to ensure that all stages are<br />
checked where there is arisk of introduction of contaminants. The samples must be<br />
representative of the process stream they are taken from. It follows that the sampling<br />
devices must be fit for this purpose. Analysis of the samples may use classical<br />
microbiological techniques, in other words, inoculation into a suitable medium,<br />
incubation <strong>and</strong> scoring for growth. This `classical' approach is a valuable aid to<br />
validating the microbiological integrity of the process. Several selective media are in<br />
common usage, which have been developed specifically for the isolation <strong>and</strong><br />
identification of brewery contaminants. Several days are usually required before aresult<br />
is obtained, therefore the data is `historical' <strong>and</strong> cannot be used for immediate hygiene<br />
control. Nevertheless, routine microbiological testing is valuable when used for trend<br />
analysis.<br />
Microbiological testing of some samples must cope with the presence of production<br />
yeast strains. Methods for the identification <strong>and</strong> differentiation of brewing strains are<br />
described in Chapter 13 (Section 13.10). Several selective media have been devised<br />
which allow the identification of bacteria <strong>and</strong> non-brewing, so-called wild yeast<br />
contaminants in the presence of high concentrations of brewing yeast. Validation of<br />
cleaning cannot be usefully checked with conventional microbiological methods since<br />
these are too slow for production requirements. Rapid procedures have been developed to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
meet this need. These have been designed to detect instantly the presence of microorganisms<br />
<strong>and</strong> soiling.<br />
17.2 The microbiological threat to the brewing process<br />
Micro-organisms may exert adverse effects on the brewing process both directly <strong>and</strong><br />
indirectly. The direct effects are the obvious ones of contamination of wort or beer with<br />
foreign organisms. Oxygenated wort represents acomparatively rich source of nutrients<br />
capable of supporting the growth of awide range of micro-organisms. The presence of<br />
hops is advantageous since trans-humulone, ( )-humulone <strong>and</strong> colupulone are inhibitory<br />
to many bacteria by virtue of their ability to act as ionophores (Verzele, 1986).<br />
Nevertheless, many micro-organisms, including yeast, are capable of growth in their<br />
presence. The effects of contamination range from comparatively minor changes in beer<br />
flavour <strong>and</strong> fermentation performance through to gross flavour defects <strong>and</strong> superattenuation<br />
of worts. Once pitched the wort is, to some extent, protected by the yeast<br />
since many contaminants, if present at low levels, are not able to compete. Beer is a<br />
comparatively poor growth medium. The nutrients are limited to the small residue that<br />
remains after fermentation is completed. Beer is arelatively hostile environment to many<br />
micro-organisms. The antiseptic properties of hop compounds are augmented by ethanol.<br />
Low redox <strong>and</strong> acid pH provide additional protection against many potential spoilage<br />
organisms. Ethanol is apowerful inhibitor of microbial growth. Low <strong>and</strong> zero alcohol<br />
beers have a much increased susceptibility to spoilage compared to their alcoholic<br />
counterparts.<br />
Several bacterial <strong>and</strong> some yeast species are capable of growth in beer. This can cause<br />
the formation of hazes, surface pellicles <strong>and</strong> many undesirable changes in beer flavour<br />
<strong>and</strong> aroma. The outward symptoms of these infections have been long recognized <strong>and</strong><br />
many are characterized as `diseases' of beers. These are usually descriptive of the<br />
changes in flavour <strong>and</strong> appearance.<br />
Micro-organisms exert indirect undesirable effects on brewing in three ways. Firstly,<br />
growth on raw materials can produce undesirable changes such that the materials do not<br />
behave normally. Secondly, the growth of contaminants on raw materials can generate<br />
microbial metabolites, which can persist into the brewing process <strong>and</strong> exert deleterious<br />
effects. Thirdly, very heavily contaminated raw materials can introduce microbial<br />
biomass that persists into green beer. Although dead, the cells can cause beer filtration<br />
problems <strong>and</strong> even beer hazes if filtration is deficient.<br />
From a microbiological st<strong>and</strong>point, the brewing process is divisible into the steps<br />
leading up to wort production followed by those that include fermentation <strong>and</strong> subsequent<br />
beer processing. The copper boil separates these two parts. This process step serves many<br />
functions, one of which is to sterilize wort (Chapters 9, 10). It follows that some<br />
microbiological contamination can be tolerated in the process steps preceding the copper<br />
boil,although withthe caveat thatthe raw materials must bewithinspecification. The steps<br />
after the copper boil are those in which the risk of contamination is highest <strong>and</strong> where the<br />
greatest caution must be exercised. Any brewing raw material that is capable of supporting<br />
microbial growth has the potential to produce unwanted metabolites that can persist<br />
through the brewing process <strong>and</strong> produce adverse effects. Water is aspecial case in that<br />
microbial metabolites can be introduced even though the organism does not have direct<br />
contact with any other brewing materials. To counteract this threat all water, especially<br />
from wells, should be carbon filtered to remove organic contaminants (Chapter 3).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Sugar syrups may become tainted as aresult of the growth of osmophilic moulds <strong>and</strong><br />
yeasts (Chapter 2). Growth on the surface of the syrup can occur where inappropriate<br />
storage conditions allow the formation of condensation resulting in alocalized sugar<br />
dilution <strong>and</strong> lower water activity. Growth of contaminants may proceed since the<br />
inhibitory effects of high osmotic potential are reduced. This should be guarded against<br />
by storage of syrups at an appropriate temperature <strong>and</strong> preferably under an inert gas such<br />
as carbon dioxide or nitrogen.<br />
Malt <strong>and</strong> adjuncts derived from cereals present the greatest threat. Poor control of the<br />
steps involved in the manufacture of these ingredients can result in mould growth.<br />
Species ofmoulds from generasuch as,Alternaria,Aspergillus, Cladosporium, Fusarium<br />
<strong>and</strong> Rhizopus have all been reported to produce adverse effects (Flannigan, 1999).<br />
Excessive mould growth produces many metabolites that can produce off-flavours <strong>and</strong><br />
aromas in beers. Terms such as molasses, stale, burned <strong>and</strong> winey have all been used to<br />
describe the effects. In addition, changes in colour may also occur. The metabolites<br />
producing these effects also result in increased nitrogen levels in worts <strong>and</strong> beers. In<br />
extreme cases, beer hazes may be generated.<br />
The most widely recognized defect ascribed to the growth of mould on malts is that of<br />
gushing. This phenomenon occurs in bottled beers where on broaching there is asudden<br />
loss of carbon dioxide with concomitant uncontrolled foaming. Studies have demonstrated<br />
that culture filtrates of several moulds, especially Fusarium spp, were capable of<br />
inducing gushing when added to beers (Amaha et al., 1974; Kitabatake <strong>and</strong> Amaha,<br />
1974). Small polypeptides have been isolated that are apparently responsible for the<br />
phenomenon. In one case aconcentration as low as 0.05ppm was sufficient to produce<br />
the effect (Kitabatake <strong>and</strong> Amaha, 1974). Moulds capable of producing gushing-inducing<br />
metabolites are commonly those that also produce mycotoxins. Indeed, some, but not all,<br />
mycotoxins have been shown to be capable of inducing gushing. More than 200 distinct<br />
mycotoxins have been isolated from various fungi. They appear to function as facilitators<br />
of fungal pathogenesis. The most common are the trichothecenes of which around 150<br />
have been recognized. Chemically, they are tetracyclic sesquiterpenes, the most common<br />
being nivalenol, deoxynivalenol <strong>and</strong> T-2 toxin. All are potent inhibitors of protein<br />
synthesis <strong>and</strong> possibly they disrupt membrane function. Trichothecenes are heat stable<br />
<strong>and</strong> therefore capable of surviving through the wort boil. They are toxic to humans <strong>and</strong><br />
animals. Purely from abrewing st<strong>and</strong>point, if present at high concentration they inhibit<br />
yeast growth. It has been suggested that they could in some circumstances be acause of<br />
sticking fermentation (Boeira et al., 1999a,b).<br />
After wort boiling the microbiological integrity of the process is dependent upon good<br />
hygienic <strong>practice</strong>. The efficiency of CIP systems is of paramount importance to ensure<br />
that contamination is not introduced by unclean plant. Uninoculated wort, either in<br />
fermenter or propagater is at the greatest risk. This must be pitched as soon as possible<br />
afteritsproductioninordertominimizetheriskofinfection.Althoughbeerisarelatively<br />
poor substrate, asubstantial range of micro-organisms are capable of growth in it. It<br />
follows, therefore, that after wort, bright beer represents the second most vulnerable<br />
material. Entry of microbial contaminants can occur at any stage where liquid or gaseous<br />
additions are made to primary process streams. Some of these are illustrated in Fig. 17.1.<br />
At the end of the process, the beer must be packaged in a way that renders it<br />
microbiologically stable throughout its expected shelf-life. In the case of small-pack<br />
products, in bottle or can, the most common option is to use a tunnel pasteurization<br />
process. The temperature <strong>and</strong> contact time must be controlled to ensure that each<br />
individual package receives the desired heat treatment. This requires careful design <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Laboratory<br />
propagation<br />
Brewery<br />
propagation<br />
Break<br />
down<br />
liquor<br />
O2/air<br />
Conditioning tank<br />
Filtration<br />
Bright beer tank<br />
Clean fill Sterile fill<br />
Tunnel<br />
pasteurizer<br />
Small pack<br />
beers<br />
Malt Liquor<br />
Hopped wort<br />
Wort collection Additives<br />
Fermentation<br />
Green beer<br />
Bulk beer<br />
tanking<br />
Flash<br />
pasteurizer<br />
Acid<br />
washing<br />
Yeast cropping<br />
Racking<br />
tank<br />
Cask<br />
Keg Dispense<br />
Pitching<br />
yeast<br />
Primings<br />
Finings<br />
Other additives<br />
Fig. 17.1 Outline of the complete brewing process indicating steps in which there is a potential for<br />
microbiological contamination.<br />
operation in the event of a packaging line stoppage. It is common <strong>practice</strong> in this situation<br />
to cool the pasteurizer to safeguard product already inside against heat degradation due to<br />
over-pasteurization. When product flow recommences, it is essential to delay forward<br />
movement of product within the pasteurizer until correct operating temperatures are<br />
attained.<br />
For certain small-pack beers where the container cannot withst<strong>and</strong> heat <strong>and</strong> for keg<br />
beers the product is pasteurized in-line. This process is as efficacious as tunnel<br />
pasteurization, however it introduces extra risks. Thus, contamination is possible from the<br />
container or from the plant situated between the pasteurizer <strong>and</strong> package. Special<br />
precautions must be taken to ensure that this does not occur. Some flavour degradation is<br />
inevitable whenever beer is heated. To eliminate this it is becoming more common to<br />
package aseptically. In this case the in-line pasteurization step is replaced by sterile<br />
filtration. The equipment must be scrupulously clean <strong>and</strong> operated to the highest<br />
st<strong>and</strong>ards of hygiene.<br />
In the case of draught beers, the possibility of microbiological spoilage extends<br />
beyond the brewery <strong>and</strong> into retail establishments. Beers in cask are to some extent<br />
protected by the endogenous flora of brewing yeast. Nevertheless, casks are vented to the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
atmosphere for dispense <strong>and</strong> therefore open to the entry of contaminants. In awellmanaged<br />
cellar this should not occur. Dispense systems for cask <strong>and</strong> keg beers are<br />
possible routes for contamination. In particular, the possibility of biofilm development<br />
must be guarded against by the use of appropriate cleaning regimes. It is essential to<br />
consider the microbiological implications of introducing new products, plant <strong>and</strong><br />
processes into the brewery. Introduction of anew raw material has the potential toimport<br />
awhole new spectrum of microbial contaminants, which have not been encountered<br />
hitherto. This is of special note where the production of beverages other than beers is<br />
introduced into abrewery. For example, the use of non-sterile fruit concentrates, which<br />
are ingredients in flavoured alcoholic beverages. If these are used in the same plant as<br />
beer, great caution should be exercised to ensure that the coexistence of these distinct<br />
productstreamsis microbiologicallyrobust.Inparticular,it should be realized that media<br />
designed for the detection of typical brewery contaminants may not be suitable for nonbrewing<br />
micro-organisms.<br />
Itisimportanttoconsideranymicrobiologicalimplicationswheremodificationstothe<br />
brewingprocessaremade.Amove from pasteurization tosterile filling heightensthe risk<br />
of microbiological failure <strong>and</strong> the precautions to prevent this need to be made<br />
correspondingly more stringent. However, in some instances apparently unrelated<br />
changes can have unexpected consequences. For example, on quality grounds there has<br />
been a gradual tightening in specification regarding maximum dissolved oxygen<br />
concentrations in product both in process <strong>and</strong> package. This decreases the overall risk of<br />
spoilage by preventing the growth of obligate aerobes. On the other h<strong>and</strong>, it provides a<br />
better selective medium for obligate anaerobes. In fact, bacterial infections of beer by<br />
anaerobes such as Pectinatus have been recorded only in relatively recent times <strong>and</strong> are<br />
taken to reflect the gradual reduction of in-process oxygen exposure (Section 17.3.3).<br />
17.3 Beer spoilage micro-organisms<br />
Beer is liable to spoilage by arange of micro-organisms, both bacteria <strong>and</strong> yeasts.<br />
Spoilage results in the formation of hazes, undesirable flavours <strong>and</strong> aromas. Although<br />
beerisrenderedunpalatable,thegrowthofcontaminantsdoesnotgenerallyleadtohealth<br />
risks. There are some exceptions to this general statement, as discussed subsequently,<br />
however pathogenic micro-organisms do not survive in beer.<br />
17.3.1 Detection of brewery microbial contaminants<br />
Routine microbiological testing in the brewery is usually restricted to enumerating<br />
populations. In many situations, no contamination whatsoever should be detected. In<br />
other cases, some contamination is inevitable. Maintaining arecord of the numbers of<br />
micro-organisms detected provides auseful method of assessing the general cleanliness<br />
of the brewery environment, the robustness of cleaning regimes <strong>and</strong> the microbiological<br />
integrity of the process. Methods must be capable of detecting low concentrations of<br />
contaminants in isolation (as in bright beer) or in the presence of high concentrations of<br />
othermicro-organisms(asindetectionoflowlevelsofbacteriainpitchingyeastslurries).<br />
In traditional <strong>practice</strong>, microbial contamination isdetected by taking asuitable sample<br />
from the brewery <strong>and</strong> inoculating it into an appropriate solid or liquid microbiological<br />
medium. Appropriate sampling devices <strong>and</strong> procedures have been developed for the<br />
routinetestingofprocess liquids,gases <strong>and</strong>surfaces.These are describedinSection 17.5.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
After incubation, to allow any micro-organisms togrow to adetectable concentration, the<br />
cultures are examined for the presence or absence of growth. Growth can be detected via<br />
the visible formation of hazes in liquids or as colonies on solid media. Many selective<br />
media are used that contain components that allow the growth of specific strains <strong>and</strong> not<br />
others. Descriptions of common microbiological media used in brewing are provided in<br />
Section 12.3.5.<br />
Growth can be quantified by performing cell counts. This may be via direct<br />
microscopic enumeration using a counting chamber. Microscopic examination of<br />
contaminants provides a useful method for preliminary identification. More commonly,<br />
cell concentrations are determined by making colony counts where serial dilutions of the<br />
test sample are streaked out into solid media or collected on a membrane which is then<br />
placed on solid medium. After incubation under appropriate conditions, any colonies that<br />
arise are assumed to have formed from single cells. Therefore, the colony count is<br />
directly proportional to the cell concentration in the original sample.<br />
Conventional microbiological techniques will continue to have a place in brewery<br />
laboratories because they are relatively inexpensive <strong>and</strong> do not require sophisticated<br />
apparatus. They suffer the drawback of slowness <strong>and</strong> produce data that is of historical<br />
interest only. More rapid techniques are needed for validating the microbiological<br />
integrity of the brewing process in real time. Rapid methods are of two general types<br />
(reviewed by Russell <strong>and</strong> Dowhanick, 1999). Firstly, those that require a growth stage<br />
before the organisms can be detected. Secondly, those able to detect very low levels of<br />
contamination in samples without the need for pre-treatment. All rapid detection methods<br />
rely on three general principles, used alone or in combination, for their operation. These<br />
are pre-concentration, low threshold of detection <strong>and</strong> specificity. Some rapid methods,<br />
particularly those with a high degree of specificity, combine elements of detection <strong>and</strong><br />
identification. These are described below.<br />
Low levels of contaminants in samples containing little extraneous solid material can<br />
be concentrated by filtration through a sterile membrane filter. The approach can be<br />
applied to both liquids <strong>and</strong> process gases. The membrane must be sufficiently porous to<br />
allow throughput of a large sample volume but have a pore size small enough to retain<br />
bacteria. Typically, 0.22 or 0.45 m membrane filters are used for this purpose. The<br />
membrane is transferred to a Petri dish containing a suitable solid nutrient medium,<br />
incubated <strong>and</strong> examined for growth. The procedure can be made rapid by examining the<br />
membrane under a microscope <strong>and</strong> looking for micro-colonies. Commonly, membranes<br />
are stained to aid visualization of the colonies. Preferably fluorescent stains are used that<br />
are incorporated into the medium. This approach has the advantage that viable <strong>and</strong> nonviable<br />
cells can be differentiated <strong>and</strong> cells can be recovered for further analysis. The<br />
micro-colony method can produce a result within approximately 24 hours, thus saving<br />
several days compared to traditional techniques.<br />
The direct epifluorescent filter technique (DEFT) uses a membrane filtration step to<br />
concentrate low concentrations of contaminants. No pre-growth stage is required because<br />
any cells trapped on the membrane are visualized by microscopic examination after<br />
staining with a fluorescent dye, usually acridine orange. This dye binds to single-str<strong>and</strong>ed<br />
RNA molecules, which are plentiful in viable cells, <strong>and</strong> produces orange fluorescent<br />
cells. Dead cells are deficient in single str<strong>and</strong>ed RNA <strong>and</strong> these stain green due to the<br />
natural fluorescence of double-str<strong>and</strong>ed RNA. Other dyes have also been used such as<br />
berberine sulphate (in conjunction with acridine orange), aniline blue, Viablue <strong>and</strong> some<br />
tetrazolium salts. Reportedly, these produce improved staining reactions, which make<br />
easier differentiation between microbial cells, inanimate debris <strong>and</strong> the membrane. The<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
procedure can be automated using computer-assisted image analysis. Results can be<br />
obtained within 30 minutes. The sensitivity of the approach is such that a single microbial<br />
cell can be detected on a membrane.<br />
Procedures that rely on pre-incubation <strong>and</strong> a low threshold of detection use indirect<br />
methods for detecting microbial growth. In clear liquid media the presence of microbial<br />
cells can be detected using turbidometry or spectrophotometry. Using appropriate<br />
apparatus changes can be detected before visual turbidity becomes apparent. The growth<br />
of micro-organisms is an exothermic process. Apparatus has been developed that detects<br />
growth-related exothermy as an electrical current via the intermediary of a thermocouple.<br />
Impedometric devices detect changes in electrical properties of the medium brought<br />
about by microbial growth. As growth proceeds, the uptake of nutrients <strong>and</strong> the formation<br />
of extracellular ionic metabolites result in a change in the conductance <strong>and</strong> capacitance of<br />
the medium. Commercial apparatus has been developed capable of measuring these<br />
changes <strong>and</strong> thereby rapidly detecting microbial growth (for a review, see Fleet, 1992).<br />
The time required to reach the threshold of detection of microbial growth using these<br />
methods is dependent upon the size of the inoculum <strong>and</strong> the growth rate of the organism<br />
therefore the time required to obtain a result is dependent on the nature of the sample <strong>and</strong><br />
the concentration <strong>and</strong> identity of the contaminants. The procedures can be used in forcing<br />
tests where the undiluted sample is assessed, for example, analysis of bright beer.<br />
Alternatively, samples may be inoculated into a suitable growth medium. The presence of<br />
low levels of bacterial contamination in the presence of high concentrations of yeast, for<br />
example in pitching slurries, can be accommodated by the incorporation of<br />
cycloheximide to inhibit yeast growth. Compared to traditional microbiological<br />
techniques the rapid procedures use expensive <strong>and</strong> relatively sophisticated apparatus.<br />
A decision must be made as to whether or not the size of the brewing operation merits the<br />
required capital investment.<br />
All living organisms contain adenosine triphosphate (ATP). Therefore, the presence of<br />
ATP is indicative of biological activity. ATP is easily detected using the phenomenon of<br />
bioluminescence (Simpson, 1999). The firefly, Photinus pyralis contains an enzyme,<br />
luciferase. In the presence of oxygen <strong>and</strong> ATP, luciferase catalyses a reaction in which a<br />
substrate termed luciferin (6-hydroxybenzothiazole) is oxidized to oxyluciferin. During<br />
the reaction ATP is hydrolysed to AMP. For each mole of ATP hydrolysed, a photon of<br />
light with an emission maximum of 532 nm is released.<br />
ATP ‡ luciferin ‡ O2 ! AMP ‡ oxyluciferin ‡ CO2 ‡ PPi ‡ light<br />
The bioluminescence reaction has been utilized for many years as the basis of an ATP<br />
determination. It is now routinely applied to the detection of ATP in industrial<br />
environments as a means of validating cleaning regimes (Ogden, 1993). Early devices<br />
relied on sampling using a sterile swab followed by rinsing into a cuvette, addition of<br />
appropriate reagents <strong>and</strong> insertion into a bioluminometer to obtain a reading. These<br />
approaches needed skilled laboratory personnel. More modern versions utilize all-in-one<br />
arrangements, where sampling device <strong>and</strong> reagents are incorporated into a single unit.<br />
Samples may be taken in the form of swabs to check the cleanliness of surfaces or as<br />
liquid taken from terminal rinse water. The measuring unit usually includes a data capture<br />
system to facilitate the maintenance of records. Operation does not require any skill other<br />
than the ability to follow simple instructions.<br />
Several commercial ATP bioluminometers are now available. Unfortunately, there has<br />
been no agreement to st<strong>and</strong>ardize the results of the measurement. Arbitrary values of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ioluminescence, termed relative light units (RLU), are used. The output from individual<br />
instruments differs for a constant ATP concentration. It is necessary, therefore, to<br />
calibrate <strong>and</strong> set st<strong>and</strong>ards for each instrument. Typically, a range of target values in<br />
RLUs is established for each location under test. The lower limit is classed as a pass, an<br />
intermediate value indicates caution <strong>and</strong> should prompt a check of the CIP system. A<br />
value above an upper limit defines a fail <strong>and</strong> indicates the need for a check of the CIP<br />
system <strong>and</strong> repeat clean.<br />
Correlation between luminescence <strong>and</strong> microbiological counts is usually poor because<br />
ATP concentration varies widely between different species <strong>and</strong> the same species in<br />
different physiological states. In general, yeast cells contain around 100 times more ATP<br />
than bacterial cells (Hysert et al., 1976). In brewing the situation is more confusing since<br />
beer also contains significant <strong>and</strong> very variable levels of ATP. Thus, Simpson et al.<br />
(1989) reported mean levels of ATP in beer of 5 nM but with a range of 0.01 to 100 nM.<br />
For this reason, bioluminometry cannot differentiate between soil <strong>and</strong> microbial<br />
contamination. It also follows that it is not as sensitive as some direct microbiological<br />
analytical techniques. The DEFT technique is capable of detecting a single cell on a<br />
membrane. ATP bioluminescence apparatus designed for hygiene testing can detect<br />
between 5 <strong>and</strong> 250 cells (Boulton <strong>and</strong> Quain, 2001). Despite the issues regarding<br />
sensitivity it is the method of choice for validating cleaning regimes both in the brewery<br />
<strong>and</strong> for dispense lines in licensed premises.<br />
Refinements of the ATP bioluminesence method have allowed the detection of<br />
individual microbial cells in beer. For example, the Sapporo Breweries in Japan have<br />
developed apparatus termed the MicroStar RMDS-SPS (Rapid Microbe Detection<br />
System ± Sapporo Special) in which beer samples are passed through a membrane filter.<br />
Reagents are sprayed directly onto the membrane <strong>and</strong> the bioluminescence due to any<br />
trapped micro-organisms detected by a photomultiplier linked to a computer for data<br />
collection. The apparatus can be used to detect very low concentrations of beer spoilage<br />
bacteria by pre-incubation of the membrane on a suitable nutrient medium for two days,<br />
prior to performing the ATP analysis (Takahashi et al., 1999a).<br />
17.3.2 Identification of brewery bacteria<br />
Even in the best-managed brewery occasional microbiological failures do occur. In these<br />
instances, identifying the organism can be a valuable aid in tracing the nature of the<br />
process failure. Identification may be made according to taxonomic principles. In<br />
addition, it may be of equal value to make identification based on more pragmatic<br />
grounds. For example, differentiation of production <strong>and</strong> wild yeast strains. Bacteria<br />
encountered during brewing are classified <strong>and</strong> identified using classical microbiological<br />
techniques such as cellular morphology, possession or lack of motility, colonial<br />
morphology when cultivated on solid media <strong>and</strong> biochemical properties. Bacterial<br />
morphologies are classified as being rod-shaped (bacilli) or spherical (cocci). The size of<br />
bacilli <strong>and</strong> cocci <strong>and</strong> whether or not the cells are borne singly, in chains or clusters are all<br />
diagnostic of individual species. The examination of cellular morphology is aided by the<br />
use of biological stains. These assist with visualizing cells for microscopic examination.<br />
The response of individual species to some stains is of taxonomic significance. The<br />
most widely used of these is the Gram stain. Gram positive bacteria are stained purple by<br />
treatment of a heat fixed smear with the dye, crystal violet. The procedure involves<br />
subsequent steps in which the smear is treated with a solution of iodine <strong>and</strong> potassium<br />
iodide (Gram's iodine) followed by washing with ethanol <strong>and</strong> counter-staining with a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
pink dye,safranin.Gram negative cells are stained pink becausethe ethanol step removes<br />
the complexofcrystalviolet<strong>and</strong>Gram'siodine from fixedcells. Thedifferentialstaining<br />
response of the two bacterial groups is due to differences in the structures of the cell<br />
walls. All bacterial cell walls contain peptidoglycan as astructural component. In Gram<br />
positive bacteria this forms amuch thicker layer compared to Gram negative types. In<br />
consequence, in the former the complex of crystal violet <strong>and</strong> Gram's iodine remains<br />
trapped following treatment with ethanol.<br />
The biochemical properties of bacteria can be assessed using anumber of tests. Some<br />
of these are summarized in Table 17.1. These tests determine fundamental aspects of the<br />
physiology of individual strains such as the ability to utilize various sources of carbon<br />
<strong>and</strong> nitrogen <strong>and</strong> to grow aerobically or anaerobically. Other procedures look for the<br />
possession of specific enzymes. For example, bacteria that possess catalase are able to<br />
decompose exogenous hydrogen peroxide. The ability to produce specific metabolic end<br />
products can be probed by anumber of tests. These usually take the form of preparing a<br />
culture of the bacteria under investigation. The presence of aparticular metabolite is<br />
confirmed by the addition to the culture of reagents that bring about acolour change.<br />
Commonly, these tests detect the formation of metabolites that would cause recognizable<br />
defects in infected beers.<br />
Several techniques have been developed that allow the rapid identification of bacteria.<br />
These use the same principles as those methods used to differentiate yeast strains<br />
(described in Chapter 13). For completeness, abrief overview of the most promising<br />
methods is given in Table 17.2. Acomprehensive review may be found in Gutteridge <strong>and</strong><br />
Priest (1999). Undoubtedly, conventional microbiological techniques in conjunction with<br />
hygiene testing via bioluminescence will continue to be the system of choice of quality<br />
control for many brewers. However, rapid methods of detection <strong>and</strong> identification are<br />
likely to become increasingly important as brewers move from quality control to quality<br />
assurance.Thisstrategyrequiresthatspecificspoilageorganismsareidentifiedassoonas<br />
possible since they pose the greatest potential threat. This is all the more important since<br />
the trend towards packaging of beers aseptically without pasteurization seems likely to<br />
grow.<br />
TheapplicationoftherapidtechniquesoutlinedinTable17.2tobeerspoilagebacteria<br />
has been reported. For example, use of DNA polymerase technology (Dimichele <strong>and</strong><br />
Llewis, 1993; Tsuchiya et al., 1992a,b), use of membrane filtration <strong>and</strong> automatic<br />
detection of bacteria with specific fluorochromes <strong>and</strong> image analysis (Yasui <strong>and</strong> Yoda,<br />
1997), electrophoretic characterization of lactate dehydrogenases of Lactobacillus brevis<br />
(Takahashi et al., 1999b), <strong>and</strong> identification of lactic acid bacteria using monoclonal<br />
antibodies (Yasui <strong>and</strong> Yoda, 1997).<br />
17.3.3 Gram negative beer spoiling bacteria<br />
Abrief description of the Gram negative bacteria associated with beer spoilage <strong>and</strong> the<br />
defects produced by their growth is given in Table 17.3 (review, Fleet, 1992). The stages<br />
in the brewing process at which these bacteria exert their effects <strong>and</strong> the defects that are<br />
produced are dependent upon the physiological capabilities of the organisms. The<br />
descriptions of the bacterial species in Table 17.3 provide an immediate indication of the<br />
stages in brewing where the results of their spoilage may become evident. Thus, acetic<br />
acid bacteria are obligate aerobes <strong>and</strong> produce acetic acid from ethanol. The<br />
concentration of ethanol that may be tolerated varies between strains. In early work,<br />
Shimwell (1936) reported that none could grow at ethanol concentrations greater than 6%<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.1 Traditional microbiological techniques used in the identification of brewery bacteria<br />
Test Basis<br />
1. Carbon source utilization Score for growth on minimal medium supplemented with various sole carbon sources.<br />
2. Nitrogen source utilization Score for growth on minimal medium supplemented with various sole nitrogen sources.<br />
3. Production of acid <strong>and</strong>/<br />
or gas<br />
As for (1), but medium supplemented with methyl red, formation of a red colour indicates the formation of acid. Incorporation of<br />
a small inverted (Durham) tube in the medium allows the detection of the formation of gas (bubble formation).<br />
4. Catalase test Pour solution of hydrogen peroxide onto surface of slope culture. Catalase positive bacteria produce copious frothing.<br />
5. Indole test Ethanolic solution of dimethylamidobenzaldehyde in presence of HCl <strong>and</strong> potassium persulphate added to peptone water culture.<br />
The presence of indole is indicated by the formation of a red coloration.<br />
6. Voges-Proskauer (VP) test The presence of acetoin <strong>and</strong> diacetyl in peptone water culture indicated by the formation of a red coloration following the<br />
addition of -naphthol <strong>and</strong> creatine.<br />
7. Formation of hydrogen<br />
sulphide<br />
Bacterial (<strong>and</strong> yeast) colonies develop black coloration when grown on nutrient medium <strong>and</strong> then overlaid with paper soaked in<br />
lead acetate solution.<br />
8. Nitrate reduction Inoculate a broth culture supplemented with 0.1% potassium nitrate. The presence of nitrite is detected by the formation of a pink<br />
colour following the addition of a solution of -naphthylamine containing acetic <strong>and</strong> sulphanilic acids.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.2 Rapid methods for the identification of brewery bacteria (Boulton <strong>and</strong> Quain, 2001)<br />
Test Principle<br />
1. Genomic analysis Hybridization of unknown genome with DNA or RNA probe(s) from known organism.<br />
2. Proteomic analysis Extraction of proteome <strong>and</strong> analysis via electrophoresis. Identification via comparison with electrophoretograms made from the<br />
proteomes of known organisms.<br />
Analysis of extracted proteome <strong>and</strong> identification of selected proteins via binding to specific labelled antibodies.<br />
Whole cell pyrolysis under inert atmosphere <strong>and</strong> analysis of sub-cellular fragments via gas chromatography (Py-GC) or mass<br />
spectroscopy (Py-MS).<br />
3. Analysis of cellular<br />
components<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Whole cell extraction of fatty acids <strong>and</strong> identification from GC profiles of fatty acid methyl esters (FAME).<br />
Analysis of whole cells via Fourier transform infra-red spectroscopy (FT IR).<br />
Analysis of whole cells via ultraviolet resonance Raman spectroscopy (UV RS)
v/v. Subsequently, other workers have reported that some strains can grow in ethanol<br />
concentrations up to 10% v/v (De Ley et al., 1984) <strong>and</strong> Gluconobacter oxydans can<br />
survive 13% v/v (Magnus et al., 1986). Some strains can grow under micro-aerophilic<br />
conditions.Theyaretolerantofethanol,hopresins<strong>and</strong>lowpH.Typically,theyspoilbeer<br />
where some oxygen is present, as might be the case in licensed premises where air is<br />
allowedtoentercasks.Beerlines<strong>and</strong>dispenseequipmentarefrequentlycontaminatedby<br />
acetic acid bacteria. Spoilage becomes evident in the form of surface pellicles, turbidity<br />
<strong>and</strong> ropiness. The latter refers to the formation of extracellular polysaccharide material,<br />
which can be seen suspended as slime in the infected beer. Infected beer becomes acid<br />
<strong>and</strong> off-flavours develop. Acetic acid bacteria are ubiquitous in brewery <strong>and</strong> licensed<br />
premises. However, they should be easily controlled by the use of appropriate hygiene<br />
regimes. In particular, dispense systems must be kept scrupulously clean. The best<br />
safeguard against acetic acid bacterial infection of beers is to eliminate oxygen.<br />
Members of the Enterobacteriaceae associated with spoilage (Obesumbacterium,<br />
Rahnella, Citrobacter <strong>and</strong> Klebsiella) are related to those such as Escherichia coli that<br />
are commonly found in the gut of mammals. For this reason they are referred to as<br />
coliforms. It should be stressed that none of those that are found as contaminants in the<br />
brewing process are pathogens. They are facultative anaerobes. In other words, they are<br />
capable of growth under both aerobic <strong>and</strong> anaerobic conditions. Inthe context of brewing<br />
this increases the number of locations where spoilage can occur. However, other<br />
metabolic constraints limit the niches that these bacteria are able to occupy. All of the<br />
species of beer spoilage Enterobacteriaceae are tolerant of hop resins <strong>and</strong> they can<br />
ferment arange of sugars but cannot utilize ethanol. In consequence they are wort<br />
contaminants <strong>and</strong> capable of exerting deleterious effects during fermentation. They do<br />
not spoil beers.<br />
Members of the Citrobacter <strong>and</strong> Klebsiella genera are sensitive to ethanol <strong>and</strong> do not<br />
survive beyond the end of fermentation. Rahnella <strong>and</strong> Obesumbacterium are more<br />
tolerant of ethanol but nevertheless do not survive very high-gravity fermentations. In<br />
lower-gravity fermentations they can persist <strong>and</strong> indeed grow with the yeast. Rahnella<br />
<strong>and</strong> especially Obesumbacterium can be cropped with the yeast, survive during storage<br />
<strong>and</strong>, if steps are not taken to remove them, they can then infect future fermentations.<br />
Growth of coliforms on wort during fermentation produces avariety of tastes <strong>and</strong> aromas<br />
ranging from sweet/honey/fruity through to vegetable/faecal. Amultitude of bacterial<br />
metabolites is responsible for these flavour changes. These include various esters, higher<br />
alcohols, organic acids, acetaldehyde, diacetyl, acetoin, dimethyl sulphide <strong>and</strong> dimethyl<br />
disulphide. This range of end products is areflection of the metabolic versatility of these<br />
bacteria. In addition to the common pathways for the metabolism of glucose via<br />
glycolysis <strong>and</strong> the hexose monophosphate shunt (Chapter 12, Figs 12.5 <strong>and</strong> 12.7) some<br />
Enterobacteria possess an alternative route, the Entner-Duodoroff pathway. In the latter,<br />
glucose is degraded to give 2-oxo-3-deoxy-6-phosphogluconate, which is then cleaved to<br />
form pyruvate <strong>and</strong> glyceraldehyde 3-phosphate (Fig. 17.2).<br />
Two major fermentative routes occur in different genera of the Enterobacteriaceae.<br />
These differ based on the relative formation of organic acids <strong>and</strong> acetoin plus 2,3butanediol<br />
(Fig. 17.3). In mixed acid types the major end products are organic acids <strong>and</strong><br />
only small amounts of acetoin <strong>and</strong> 2, 3-butanediol are formed. These types, which include<br />
Citrobacter <strong>and</strong> Obesumbacter, test positive with the methyl red test but negative with<br />
the Voges-Proskauer test. The non-mixed acid group, which includes Klebsiella <strong>and</strong><br />
Rahnella produce high concentrations of acetoin <strong>and</strong> 2, 3-butanediol <strong>and</strong> in consequence<br />
they test positive with the Voges-Proskauer procedure.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.3 Gram negative beer spoilage bacteria (Van Vuuren, 1999)<br />
Bacterial type Description Effects of growth in brewing process<br />
Acetic acid bacteria ± Acetobacter<br />
A. aceti<br />
A. liquefaciens<br />
A. pastorianus<br />
A. hansenii<br />
Acetic acid bacteria ± Gluconobacter<br />
G. oxydans<br />
Zymomonas<br />
Z. mobilis<br />
Obesumbacterium (Hafnia)<br />
O. proteus<br />
Citrobacter<br />
C. freundii<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Slightly curved or straight rods up to 4 m in length. Cells are<br />
pleomorphic <strong>and</strong> occur in pairs or chains. Some species are<br />
motile obligate aerobes <strong>and</strong> catalase positive. Capable of<br />
oxidizing ethanol.<br />
Similar morphology to Acetobacter. Obligate aerobes, catalase<br />
positive, ethanol is oxidized to acetic acid. Ethanol is not<br />
oxidized.<br />
Short fat rods, which occur singly, in pairs, chains or rosettes.<br />
No endospores are formed. Some species are motile others are<br />
not. They grow anaerobically but are catalase positive <strong>and</strong><br />
tolerate aerobiosis. Glucose <strong>and</strong> fructose (but not maltose) are<br />
fermented to form ethanol. The optimum growth temperature is<br />
25±30 ëC.<br />
Short, fat, pleomorphic rods. They are catalase positive <strong>and</strong><br />
ethanol tolerant. Growth in wort produces dimethyl sulphide,<br />
higher alcohols <strong>and</strong> diacetyl. Nitrate or nitrite are reduced to<br />
form carcinogenic nitrosamines.<br />
Slender straight rods occurring singly or in pairs <strong>and</strong> usually<br />
motile. Cells are catalase positive <strong>and</strong> are facultative anaerobes.<br />
Citrate is used by most but not all species. Glucose is fermented<br />
to form mixtures of organic acids (lactate, pyruvate, isocitrate<br />
<strong>and</strong> succinate). Relatively ethanol intolerant.<br />
Form hazes or pellicles in beers containing oxygen.<br />
Products of metabolism include acetic acid <strong>and</strong><br />
acetate.<br />
As for Acetobacter.<br />
Exclusive to ale breweries where spoilage causes<br />
`rotten apple' flavour due to the formation of<br />
hydrogen sulphide <strong>and</strong> acetaldehyde.<br />
Contaminant of pitching yeast which, if present,<br />
grows with yeast during fermentation <strong>and</strong> results in<br />
slow attenuation rates <strong>and</strong> high pH beer. Gives rise<br />
to fruity/parsnip off-flavours.<br />
Rare contaminant in fermentations where it causes<br />
accelerated attenuation rate <strong>and</strong> produces increased<br />
organic acids <strong>and</strong> DMS. They are killed in late<br />
fermentation by the presence of ethanol.
Enterobacter (Rahnella)<br />
R. aquatilis<br />
E. agglomerans<br />
Klebsiella<br />
K. terrigena<br />
K. oxytoca<br />
Pectinatus<br />
P. cerevisiiphilus<br />
Megasphaera<br />
M. cerevisiae<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Short squat rods, which may be motile. Glucose is fermented to<br />
produce acid <strong>and</strong> gas. Strains are positive in the VP test (Table<br />
17.2).<br />
Slender straight capsulated, non-motile rods occurring singly or<br />
in short chains. They are facultative anaerobes <strong>and</strong> ferment<br />
glucose to produce acid <strong>and</strong> gas.<br />
Very slender curved rods occurring singly or in pairs. Older<br />
cells are elongated. They are motile <strong>and</strong> obligately anaerobic.<br />
Obligately anaerobic slightly elongated<br />
non-motile <strong>and</strong> non-spore forming cocci occurring singly or in<br />
short chains. They are relatively ethanol intolerant.<br />
In the brewing process it behaves in a similar manner<br />
to Obesumbacterium <strong>and</strong> is a contaminant of<br />
pitching yeast. It is relatively intolerant to ethanol<br />
<strong>and</strong> survives more readily in top cropping ale<br />
fermentations. Abnormally high diacetyl levels are<br />
produced in contaminated worts.<br />
Ferulic acid in wort is decarboxylated to produce 4vinylguaiacol.<br />
This imparts a phenolic off-flavour to<br />
beer. The reaction is also catalysed by some wild<br />
yeasts.<br />
Contaminants of small-pack beers where oxygen<br />
levels are low. Produces hydrogen sulphide <strong>and</strong> other<br />
sulphur compounds.<br />
Spoilage is restricted to low oxygen environments<br />
where the ethanol concentration does not exceed c.<br />
4% v/v. Putrid aromas <strong>and</strong> tastes occur due to the<br />
formation of hydrogen sulphide <strong>and</strong> other<br />
sulphur-containing metabolites.
Malate<br />
2H<br />
Glucose<br />
Glucose 6-phosphate<br />
2-oxo-3-deoxy-6-phosphogluconate<br />
Glyceraldehyde<br />
3-phosphate<br />
Pyruvate<br />
Indole-negative Klebsiella strains produce phenolic off-flavours in beers as a<br />
consequence of the formation of 4-vinlyguaiacol via the decarboxylation of ferulic acid.<br />
The latter is a phenolic acid present in wort <strong>and</strong> the reaction is similar to that which is<br />
catalysed by some wild yeast infections. The increased concentration of dimethyl<br />
sulphide (DMS) associated with many Enterobacteriaceae infections does not derive<br />
from the degradation of sulphur-containing wort amino acids (Wainwright, 1972). In the<br />
case of R. aquatalis, DMS was formed via the reduction of dimethyl sulphoxide (McCaig<br />
<strong>and</strong> Morrison, 1984).<br />
ATP<br />
ADP<br />
H2O<br />
Fig. 17.2 The Entner-Duodoroff Pathway.<br />
Glucose<br />
Oxaloacetate Phosphoenolpyruvate<br />
CoA-SH<br />
Acetate<br />
CO2<br />
Lactate Pyruvate<br />
2H<br />
Acetyl-CoA<br />
4H<br />
CoA-SH<br />
Ethanol<br />
Formate<br />
CO2 + H2<br />
Pyruvate<br />
CO2<br />
CO2<br />
α-Acetolactate<br />
Acetoin<br />
2H<br />
2,3-Butanediol<br />
Fig. 17.3 Glucose metabolism by bacteria belonging to the genus Enterobacteriaceae (redrawn<br />
from Van Vuuren, 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
CO2
Of all the Gram negative beer spoilage bacteria, Obesumbacterium proteus poses the<br />
greatest risk to the brewing process as aresult of its role in the formation of Nnitrosamines<br />
(Smith, 1994). These compounds are powerful animal carcinogens. Nonvolatile<br />
N-nitrosamines (apparent total N-nitroso compounds, ANTC) are formed by<br />
reactions between wort amines <strong>and</strong> nitrite. Nitrate is present in all worts <strong>and</strong> O. proteus<br />
is capable of reducing it to nitrite thereby providing the precursor for ANTC formation.<br />
TheabilityofO.proteustogrowinwort<strong>and</strong>surviveintotheyeastcrophasnecessitated<br />
the introduction of procedures for ensuring that it is eliminated before re-pitching. The<br />
process of acid washing, in which pitching yeast is subjected to controlled acidification,<br />
accomplishes this. Disinfection of pitching yeast by acid washing relies upon the<br />
relative tolerance <strong>and</strong> sensitivity of yeast <strong>and</strong> bacteria, respectively to low pH (see<br />
Section 17.6).<br />
Zymomonas mobilis tolerates oxygen but grows under anaerobic conditions. It<br />
ferments glucose <strong>and</strong> fructose but not maltose. Unlike most of the Enterobacteriaceae it<br />
toleratesethanol<strong>and</strong>reportedlysurviveshigh-gravityfermentationsinwhich12±13%v/v<br />
ethanol are formed (Magnus et al., 1986). It has arelatively high optimum growth<br />
temperature of 25±30ëC. For this reason, it tends to be a more common spoilage<br />
bacterium in ale breweries as opposed to those fermenting lager worts at lower<br />
temperatures. Infected worts develop a characteristic rotten apple odour due to the<br />
formation of acetaldehyde. In addition, ethanol, acetic acid, lactic acid, acetoin <strong>and</strong><br />
glycerol are formed (Van Vuuren, 1999). The formation of <strong>and</strong> high tolerance to ethanol<br />
has made Zymomonas the organism of choice for many industrial alcohol production<br />
processes. This fact emphasizes the threat that the organism poses to brewing.<br />
The Gram negative cocci, Megasphaera spp. <strong>and</strong> the Gram negative rods, Pectinatus<br />
spp. are contaminants of beer. They are tolerant of hop resins but their potential for<br />
spoilage is limited by virtue of their absolute requirement for anaerobiosis. For this<br />
reason they tend to be found in finished beers. Megasphaera strains produce several<br />
organic <strong>and</strong> fatty acids, notably butyric acid <strong>and</strong> some acetic, isovaleric <strong>and</strong> valeric. In<br />
addition, hydrogen sulphide is generated (Engelmann <strong>and</strong> Weiss, 1985). Their potential<br />
for beer spoilage is restricted by their sensitivity to ethanol (>2.8% v/v) <strong>and</strong> acid pH<br />
(Haikara <strong>and</strong> Lounatmaa, 1987). It is considered that unpasteurized, low-alcohol beers<br />
aremostpronetospoilagebyMegasphaera.Nevertheless,severalweeksmayberequired<br />
before turbidity becomes evident.<br />
Pectinatus strains are also obligate anaerobes but are more tolerant of ethanol. Their<br />
presence in pitching yeast has been reported (Haikara, 1989) butinfection of beer via this<br />
routeisofvery rare occurrence.Spoilage ofbeerbyPectinatusresultsintheformationof<br />
high concentrations of hydrogen sulphide with its putrid odour <strong>and</strong> development of<br />
turbidity. Various fatty acids, especially propionic <strong>and</strong> acetic, together with some acetoin<br />
are also produced.<br />
17.3.4 Gram positive beer spoiling bacteria<br />
Gram positive bacteria associated with beer spoilage are either rods or cocci, which<br />
together are termed the lactic acid bacteria. In addition, some members of the genera<br />
Bacillus <strong>and</strong> Micrococcus have been isolated from beers, although their status as true<br />
spoilage bacteria is questionable. Nevertheless they are included for completeness (Table<br />
17.4). Apart from their reaction to the Gram stain this group of spoilage bacteria is<br />
distinguished from Gram negative types in that they are on average less resistant to the<br />
antiseptic effects of hop resins. However, this distinction is not absolute <strong>and</strong> there is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.4 Gram positive beer spoilage bacteria (Priest, 1999)<br />
Bacterial type Description Effect on brewing process<br />
Lactobacillus<br />
L. brevis<br />
L. casei<br />
L. plantarum<br />
L. fermentum<br />
L. buchneri<br />
L. delbruÈckii<br />
Pediococcus<br />
P. damnosus<br />
(syn. P. cerevisiae)<br />
P. inopinatus<br />
Bacillus<br />
B. coagulans<br />
Micrococcus<br />
M. kristinae<br />
Slender non-motile anaerobic rods that do not form<br />
endospores. They lack catalase but can tolerate oxygen <strong>and</strong><br />
low pH. Some strains are resistant to hop resins. They usually<br />
have fastidious nutritional requirements. Fermentative growth<br />
produces mainly lactic acid (homofermentative types) or<br />
mixtures of lactic acid, acetic acid, ethanol <strong>and</strong> carbon dioxide<br />
(heterofermentative types).<br />
Gram positive non-motile cocci occurring singly, in pairs or as<br />
tetrads/short chains. Originally they were known as sarcinae,<br />
although aggregates of eight cells are rare. They are catalase<br />
negative but can tolerate some oxygen <strong>and</strong> grow under<br />
microaerophilic conditions. Most strains are homofermentative<br />
<strong>and</strong> many are resistant to hop resins. They are ethanol tolerant.<br />
Large motile rods which form endospores. They are catalase<br />
positive <strong>and</strong> aerobic/facultatively anaerobic. They are<br />
thermoduric <strong>and</strong> thermophilic but sensitive to hop resins <strong>and</strong><br />
cannot grow in media with a pH lower than c. 5.0.<br />
They are catalase positive <strong>and</strong> usually obligate aerobes (M.<br />
kristinae is a facultative anaerobe). They are sensitive to acid<br />
pH <strong>and</strong> hop resins.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Produce turbidity in infected beers. Some strains produce<br />
extracellular polysaccharides, which appear as visible `ropes' in<br />
infected beer. Sour/acid off-flavours are generated.<br />
Spoilers of fermenting worts <strong>and</strong> beers where they produce<br />
hazes, acidity <strong>and</strong> high concentrations of diacetyl. Historically,<br />
the latter was referred to `sarcina sickness'.<br />
The endospores allow them to survive wort boiling. They are<br />
able to grow in hot (55±70 ëC) sweet wort where they produce<br />
lactic acid. They are inhibited by hop acids <strong>and</strong> low pH <strong>and</strong> do<br />
not cause beer spoilage.<br />
Common contaminants in breweries but their sensitivity to hop<br />
resins <strong>and</strong> intolerance of acid pH prevent beer spoilage.
significant variability. Thus, some members of the lactic acid bacteria are resistant to hop<br />
resins, whereas Micrococcus <strong>and</strong> Bacillus spp. are sensitive.<br />
The lactic acid bacteria are an important group. Apart from having the potential to<br />
spoil foods they are used industrially to make fermented dairy products such as yoghurt.<br />
Others are of clinical significance. A review of these bacteria may be found in Priest<br />
(1999). The genus Lactobacillus contains members that are genetically diverse <strong>and</strong><br />
further revision <strong>and</strong> sub-division is likely. Original classifications were based upon the<br />
mode of fermentative growth <strong>and</strong> temperature relations. Heterofermentative types were<br />
placed within the Betabacterium genus. Homofermentative types were subdivided into<br />
thermophilic strains (Thermobacterium) <strong>and</strong> mesophilic strains (Streptobacterium).<br />
These groups are still mentioned but do not have taxonomic significance.<br />
Lactococcal bacteria are classified into several genera. Historical classifications<br />
placed many of them into the Streptococci on the basis of facultative anaerobiosis <strong>and</strong><br />
homofermentative physiology. These have now been sub-divided into the Streptococcus<br />
sensu stricto, Enterococcus (indicators of faecal contamination, includes S. faecalis),<br />
Lactococcus (includes the dairy types, S. lactis), Vagococcus (motile types resembling<br />
Lactococcus). Heterofermentative cocci occurring in pairs or short chains are now<br />
classified as Leuconostoc. Homofermentative cocci that divide in two planes to produce<br />
pairs or tetrads are classified as Pediococcus.<br />
Beer spoilage by lactococcal bacteria is restricted to Pediococcus, the most common<br />
species being P. damnosus. This bacterium is found commonly as a contaminant of wort<br />
<strong>and</strong> beer. It is not found in brewing raw materials (Priest, 1999), suggesting that it is<br />
particularly well adapted to the environment of the brewery. A second species, P.<br />
inopinatus, has also been isolated from breweries but is also found in non-brewing<br />
habitats. P. inopinatus has been isolated from pitching yeast but rarely from beer. Of the<br />
two species, P. damnosus is the more resistant to hop resins <strong>and</strong> it persists through<br />
fermentation into finished beer.<br />
The potential for Lactobacillus to spoil beer is dependent upon the relative sensitivity<br />
of individual strains to hop resins. Simpson <strong>and</strong> Fern<strong>and</strong>ez (1992) determined the<br />
minimum concentration of trans-isohumulone required to inhibit the growth of 42 strains<br />
of Lactobacillus. The bacteria could be classified into three groups; sensitive types<br />
inhibited by 20 M trans-isohumulone, an intermediate group in which the minimum<br />
inhibitory concentration was 20±40 M <strong>and</strong> a third group capable of growth in the<br />
presence of up to 180 M trans-isohumulone. Only the third group could be isolated from<br />
beer. In a later paper (Simpson <strong>and</strong> Fern<strong>and</strong>ez, 1994), the same authors concluded that<br />
resistance to at least 90 M trans-isohumulone was necessary for Lactobacillus strains to<br />
qualify as beer spoilers. Resistance to hop resins is pH dependent. Simpson (1993)<br />
reported that increase in pH decreases the toxic effect of hop iso- -acids. In the author's<br />
view, an increase in pH of as little as 0.2 units could reduce the protective effect of hop<br />
resins by as much as a half.<br />
Confirmation that Lactobacillus strains are beer spoilers is notoriously difficult to<br />
demonstrate since usually they will not grow on beer. Cultivation may be facilitated by<br />
successive pre-incubation of strains in media containing increasing proportions of beer.<br />
Similar results are obtained if the beer is substituted with 45 M trans-isohumulone<br />
(Simpson <strong>and</strong> Fern<strong>and</strong>ez, 1992). This training procedure is time consuming. A more<br />
rapid approach has been developed in which MRS medium is supplemented with 20 M<br />
trans-isohumulone (Simpson amd Hammond, 1991). This medium suppresses the growth<br />
of non-beer spoiling Lactobacillus strains but permits the growth of those capable of<br />
spoilage. The ability of some strains to tolerate hop resins is probably a consequence of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the possession of a plasmid-borne gene, termed hor A. Thus, Sami et al., (1997a)<br />
demonstrated that of 61 strains containing hor A, only two could not grow in hopped<br />
beer. Similarly, only one out of 34 hor Anegative strains grew in beer. The same group<br />
hasproposedthat rapididentificationofbeer spoiling Lactobacillusstrains ispossiblevia<br />
detection of hor Ausing PCR DNA technology. From ataxonomic st<strong>and</strong>point, at least<br />
nine species of Lactobacillus have been isolated from beer (Sami et al., 1997b). Those<br />
most commonly encountered are L. brevis, L. casei, L. curvatus, L. plantarum <strong>and</strong> L.<br />
delbruÈckii.<br />
Infections of beer by Pediococci are characterized by the formation of high<br />
concentrations of diacetyl, accompanied by a reduction in yeast growth <strong>and</strong> low<br />
fermentation rates. Historically, this was called sarcina sickness, a reference to the<br />
similarity between Pediococci <strong>and</strong> true octuplets of Sarcina spp. Extracellular<br />
thixotrophic polysaccharide slimes may also be formed resulting in the formation of<br />
visible `rope'. Infections by Lactobacillus produce similar symptoms to Pediococcus.<br />
Growth in beer produces `silky turbidity'. Rope may be formed by some strains. As with<br />
Pediococcus acid is produced although the most noticeable flavour defect is the<br />
formation ofdiacetyl.Alllacticacid bacteria rely onthe small concentrations ofnutrients<br />
available in beer for their growth. The presence of priming sugars, such as sucrose or<br />
fructose, provides areadily assimilable source of sugar. In the absence of simple sugars,<br />
maltotriose <strong>and</strong> maltotetrose may be utilized by some strains. L. diastaticus, now not<br />
considered to be aseparate species <strong>and</strong> placed within L. brevis, is capable of utilizing<br />
dextrins. As its name suggests it has the potential to produce superattenuation of worts<br />
during fermentation.<br />
Lactic acid bacteria are catalase negative anaerobes. They do not possess superoxide<br />
dismutase (Archibald <strong>and</strong> Fridovich, 1981). Since they lack these st<strong>and</strong>ard mechanisms<br />
for nullifying the toxic effects of reactive oxygen radicals, it is surprising that they can<br />
tolerate exposure to oxygen. In fact, they use acombination of NADH oxidases <strong>and</strong> a<br />
pseudocatalase for removing peroxide ions (Johnston <strong>and</strong> Delwiche, 1965). Superoxide<br />
radicals are apparently scavenged by high intracellular concentrations of Mn 2+ .<br />
In homofermentative strains (Pediococcus spp., L. casei, L. plantarum <strong>and</strong> L.<br />
delbruÈckii) the major product of sugar metabolism is lactic acid. In this group sugars are<br />
metabolized via glycolysis. Pyruvate, derived from glycolysis, is reduced to lactate via<br />
the action of NADH-linked lactate dehydrogenase. Heterofermentative strains (L. brevis)<br />
produce amixture of end products, including lactate, glycerol, ethanol <strong>and</strong> acetate. These<br />
strainsutilize the phosphoketolasepathway, inwhich, glucosecatabolismproceedsvia 6phosphogluconate<br />
which, following adecarboxylation reaction, forms the pentose sugar,<br />
xylulose 5-phosphate. These reactions are those that form the initial part of the hexose<br />
monophosphate shunt (Section 12.5.2). Apart from this being an alternative to glycolysis,<br />
it provides aroute by which pentoses can be catabolized. Phosphoketolase catalyses the<br />
cleavage of xylulose 5-phosphate to yield, after the addition of another phosphate group,<br />
amolecule of triose phosphate <strong>and</strong> one of acetyl phosphate. From these intermediates<br />
glycerol, acetate, ethanol <strong>and</strong> pyruvate are formed (Fig. 17.4).<br />
The formation of diacetyl by lactic acid bacteria contamination during fermentation<br />
does not exclusively use the same pathway as that employed by brewing yeast. In yeast,<br />
diacetyl derives from the spontaneous oxidative decarboxylation of -acetolactate.<br />
Diacetyl is then reduced to acetoin <strong>and</strong> 2,3-butanediol (Section 12.10.2). In lactic acid<br />
bacteria, the same sequence of reactions can occur but some strains also possess -<br />
acetolactate decarboxylase, which produces acetoin directly, without the intermediary of<br />
diacetyl. A second route also operates in which diacetyl is synthesized directly from the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Maltose<br />
Glucose<br />
Glusose<br />
1-phosphate<br />
Glucose<br />
6-phosphate<br />
Fructose<br />
1,6-bisphosphate<br />
Glyceraldehyde<br />
3-phosphate<br />
Pyruvate<br />
Lactate<br />
6-Phosphogluconate<br />
Glyceraldehyde<br />
3-phosphate<br />
Pentoses<br />
Xylulose<br />
5-phosphate<br />
Acetyl<br />
phosphate<br />
Glycerol Pyruvate Acetaldehyde<br />
Lactate Ethanol<br />
Acetate<br />
Fig. 17.4 Homofermentative <strong>and</strong> heterofermentative sugar metabolism in lactic acid bacteria (for<br />
details see text).<br />
TPP<br />
Pyruvate α-Acetolactate<br />
Acetaldehyde-TPP<br />
activated form of acetaldehyde (acetaldehyde thiamine pyrophosphate) <strong>and</strong> acetyl-CoA<br />
(Speckman <strong>and</strong> Collins, 1973; Fig. 17.5).<br />
17.3.5 Beer spoilage yeasts<br />
Spoilage by yeasts is potentially a serious problem since many such contaminants are<br />
capable of occupying the same ecological niche as production strains. However, since<br />
they have different genotypes from the production strain their activities can produce a<br />
variety of defects in process <strong>and</strong> product. Traditionally, contaminants are referred to as<br />
wild yeasts. The concept of `wildness' is imprecise. Thus, contaminants may range from<br />
non-Saccharomyces yeast strains through to accidental mixing of production brewing<br />
strains. In the latter case, the mixing of ale <strong>and</strong> lager strains can be especially<br />
problematic. A good working definition of wild yeast is any yeast not deliberately used<br />
<strong>and</strong> under full control (Gillil<strong>and</strong>, 1971). This all-embracing definition allows for all<br />
possibilities, from the use of pure monocultures through to those rare fermentations that<br />
rely on spontaneous contamination. The similarity of wild yeasts to production strains can<br />
make them difficult to detect. Although their presence may be signalled by major changes<br />
CO2<br />
Acetyl-CoA CoA-SH<br />
Diacetyl<br />
Fig. 17.5 Pathways for the formation of diacetyl by lactic acid bacteria.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
to product <strong>and</strong> process, it is equally possible that much less apparent <strong>and</strong> subtle defects<br />
can be caused. Acid washing of pitching yeast, which reduces bacterial contamination is<br />
not effective with wild yeast. Avoidance of contamination by yeast is entirely dependent<br />
upon the maintenance of high st<strong>and</strong>ards of hygiene.<br />
Beer spoilage yeasts may be considered as Saccharomyces <strong>and</strong> non-Saccharomyces<br />
types. The Saccharomyces wild yeasts pose the greatest threat since they are most similar<br />
to production strains. Ipso facto, they have the ability to colonize the same range of<br />
habitats as production yeast strains. Their similarity to production strains makes<br />
differentiation difficult. The taxonomic classification of Saccharomyces beer spoilers is<br />
now of little practical significance in that many strains originally given the status of<br />
species, based on characteristics of relevance to spoilage, have now been assigned to S.<br />
cerevisiae. The discussion in this section will be confined to a description of the effects of<br />
contamination. Species names will be used only to provide a historical context. The<br />
effects of inadvertent mixing of production strains are difficult to predict. Some likely<br />
outcomes are changes in flavour, cropping behaviour, fining behaviour <strong>and</strong> attenuation<br />
rate <strong>and</strong> extent. These are immediately obvious changes. More subtle changes from the<br />
norm, particularly if the level of contamination is low, may be much more difficult to<br />
detect, especially if there is a gradual change in the level of contamination with<br />
successive fermentations.<br />
More dramatic defects are caused by contamination with specific Saccharomyces<br />
yeasts. Certain Saccharomyces strains <strong>and</strong> some members of the genera Kluyveromyces,<br />
Pichia <strong>and</strong> Williopsis synthesize so-called killer factors. These are toxins, also known as<br />
zymocins that have no effect on the producing strain but are rapidly lethal to other<br />
susceptible strains of the same species (Young, 1987; Magliani et al., 1997). Some<br />
zymocins are ionophores. They exert their lethal effects by disrupting the plasma<br />
membrane of target cells such that their ability to retain ions is destroyed. Others inhibit<br />
DNA synthesis.<br />
Contamination of fermentations with killer yeast has the potential for catastrophic<br />
disruption of the brewing process. They appear to be rare in breweries. In a survey of 964<br />
species, representing 28 genera, 59 were found to produce killer factors. Of these, more<br />
than half were Saccharomyces strains (Sami et al., 1997a). Most of these were laboratory<br />
haploids <strong>and</strong> only four were brewing types (Philliskirk <strong>and</strong> Young, 1975). Occasional<br />
infections with killer yeast of an early continuous primary fermentation system were<br />
reported by Maule <strong>and</strong> Thomas (1973). The problem was severe, since infection levels of<br />
less than 3% of the total yeast population were sufficient to virtually eliminate the<br />
production strain.<br />
Killer factors might be harnessed for a useful purpose. An early proposal was to<br />
genetically modify brewing strains by the introduction of a killer factor (Hammond <strong>and</strong><br />
Eckersley, 1984). The use of a `killer' brewing yeast strain would be an aid in the<br />
prevention of contamination. This option has not been pursued since brewers will not use<br />
genetic engineering. No doubt that it is an interesting approach but it would never be a<br />
substitute for good hygiene. In addition, it might be a risky strategy where several yeast<br />
strains are used within the same brewery.<br />
Most brewing strains are unable to utilize dextrins <strong>and</strong> these persist in beer where they<br />
contribute to fullness <strong>and</strong> mouth-feel. Some strains, originally classified as S. diastaticus<br />
but now placed with S. cerevisiae, possess glucoamylase <strong>and</strong> in consequence can utilize<br />
dextrins. Contamination of fermentations with diastatic yeasts leads to superattenuation<br />
of the wort <strong>and</strong> beers with abnormally low present gravity. Occasionally, diastatic yeasts<br />
have been used to produce so-called `light' beers. However, such strains commonly<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
OH<br />
OCH3<br />
CO2<br />
possess other, undesirable characteristics. The removal of dextrins is more usually<br />
accomplished by the direct addition to wort of preparations of glucoamylase<br />
(amyloglucosidase). Contamination of unpasteurized bottled beer with diastatic yeast is<br />
potentially hazardous, since abnormally high concentrations of carbon dioxide can<br />
develop with the consequent risk of bottle explosions.<br />
Many diastatic yeast strains possess a gene termed POF, an acronym for phenolic offflavour<br />
(Ryder et al., 1978). This gene encodes for the enzyme phenolic acid<br />
decarboxylase. This enzyme decarboxylates wort phenolic acids such as ferulic <strong>and</strong><br />
cinnamic acids to produce 4-vinyl guaiacol <strong>and</strong> styrene (Fig. 17.6). These compounds<br />
impart a medicinal or clove-like taste <strong>and</strong> aroma. They are an essential part of the flavour<br />
of some beers, for example, many wheat beers. In most cases the presence of these<br />
compounds is a serious defect.<br />
Many non-Saccharomyces yeasts are routinely found in breweries. Most cannot<br />
compete with brewing yeasts <strong>and</strong> hence do not usually gain a foothold within the process.<br />
Many potential contaminants are not particularly tolerant of ethanol, few are able to<br />
ferment sugars <strong>and</strong> many cannot grow under anaerobic conditions. Their relatively poor<br />
adaptation to brewing conditions means that the threat they pose is small. Spoilage by<br />
non-Saccharomyces yeast is generally restricted to some raw materials <strong>and</strong> to the aerobic<br />
phase of fermentation. More opportunities for spoilage occur in licensed premises. Ales<br />
in cask <strong>and</strong> dispense systems used for all draught beers can provide a semi-aerobic<br />
environment in which many non-Saccharomyces yeasts can grow. In this sense these<br />
yeasts are opportunistic contaminants which flourish where poor hygiene <strong>and</strong> bad<br />
<strong>practice</strong> combine to provide the conditions for growth. Scrupulous cleaning of dispense<br />
equipment <strong>and</strong> prevention of air ingress into casks by dispensing under blankets of inert<br />
gas minimize the risks.<br />
Non-Saccharomyces yeasts encountered in both the brewery <strong>and</strong> licensed premises<br />
include representatives of the following genera. Cryptococcus <strong>and</strong> Rhodotorula are<br />
commonly detected but unless conditions are grossly atypical are not able to spoil wort or<br />
beer. C<strong>and</strong>ida, Kluyveromyces, Pichia <strong>and</strong> Torulaspora are opportunistic spoilers during<br />
the aerobic phase of fermentation <strong>and</strong> in unpasteurized cask beers. Most are obligate<br />
aerobes although C<strong>and</strong>ida <strong>and</strong> Torulaspora are capable of poor growth under anaerobic<br />
OH<br />
CH CH COOH<br />
CH CH2<br />
Ferulic acid 4-Vinylguaiacol<br />
CH CHCOOH CH CH2<br />
CO2<br />
Cinnamic acid Styrene<br />
OCH3<br />
Fig. 17.6 Formation of 4-vinylguaiacol from ferulic acid <strong>and</strong> styrene from cinnamic acid by wild<br />
yeast possessing the phenolic off-flavour (POF) gene.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
conditions. Strains of Pichia which colonize cask ales maximize their opportunity for<br />
utilizing any oxygen in the gas space by forming surface films. Zygosaccharomyces,<br />
especially Z. bailii <strong>and</strong> Z. rouxii are osmotolerant strains <strong>and</strong> can cause spoilage of bulk<br />
sugar syrups. Of all the non-Saccharomyces yeasts, Brettnanomyces <strong>and</strong> Dekkera<br />
probably pose the greatest threat to unpasteurized beers. Both are able to ferment sugars<br />
to form ethanol, however oxygen stimulates fermentation, (Custers effect, Section<br />
12.5.8). Beer spoilage by these strains is characterized by the formation of high<br />
concentrations of acetic acid.<br />
17.3.6 Microbiological media <strong>and</strong> the cultivation of micro-organisms<br />
In order to cultivate micro-organisms in the laboratory it is necessary to provide<br />
favourable growth conditions <strong>and</strong> it may also be important to control conditions such that<br />
they are inimical for organisms whose growth is not desired. This requires the provision<br />
of asuitable source of nutrients, possibly the addition of growth inhibitors <strong>and</strong> control of<br />
the physical environment.<br />
The key environmental parameters are temperature <strong>and</strong> gas supply. The former is<br />
regulated by the use of thermostatically controlled microbiological incubators.<br />
Maintenance of temperature to +/ 1ëC within arange of ambient to 55ëC is usually<br />
adequate. In many instances cultivation at room temperature (18±25ëC) is sufficient. The<br />
gas supply is controlled to differentiate between aerobes <strong>and</strong> anaerobes. For aerobic<br />
cultures incubation is in the presence of air. For liquid cultures it may be necessary to<br />
improve oxygen transfer from the air to the growing micro-organisms by incubating on<br />
devices which shake the culture flask. Commonly, flask shakers <strong>and</strong> thermostatically<br />
controlled incubators are combined into asingle piece of apparatus. For larger volume<br />
liquid cultures, where shaking is not practicable or efficient, air or oxygen can be<br />
introducedbybubblinggasintothemediumthroughasinterorc<strong>and</strong>lefittedwithasterile<br />
filter. The provision of mechanical stirring further improves gas transfer rates.<br />
Maintenance ofanaerobic conditions requires specializedequipment.Typically,thisis<br />
ajar in which solid or liquid cultures are placed. Air can be removed by evacuation or by<br />
replacement with an inert gas such as nitrogen. Visual confirmation of anaerobiosis is<br />
provided by incorporation of asolution of methylene blue, which is colourless in its<br />
reduced form. Early devices relied on flushing out air with hydrogen <strong>and</strong> removing the<br />
final vestiges of oxygen via combustion using aplatinum catalyst. Modern approaches<br />
use commercially available kits that remove oxygen chemically. Typically, these are<br />
added to an anaerobic jar <strong>and</strong> activated by the addition of water. For example, one such<br />
kit contains sodium borohydride <strong>and</strong> sodium bicarbonate. Addition of water generates<br />
hydrogen <strong>and</strong> carbon dioxide, which flushes out air. For incubation of one or asmall<br />
number of plates, these kits are available as self-contained re-sealable foil pouches<br />
containing deoxygenating chemicals. Enrichment of the atmosphere within an anaerobic<br />
culture with carbon dioxide promotes the growth of many organisms, for example, lactic<br />
acid bacteria. The choice of solid or liquid medium depends upon the application. Liquid<br />
media are often used for the production of pure cultures, as with yeast propagation<br />
(Section 13.5.2). They are useful where the yield <strong>and</strong> growth rate must be maximized.<br />
Solid media are useful for characterizing microbial populations. They consist of a<br />
nutrient medium solidified with agar or gelatin. Normally the medium is held in a Petri<br />
dish (Prescott et al., 1996).<br />
Solid media serve several useful purposes. Microbiological samples can be plated out<br />
such that individual cells are separated from their neighbours. During incubation colonies<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
develop <strong>and</strong> it may be assumed that each derives from asingle cell. Counting colonies<br />
allows enumeration of the microbial population within the sample. The appearance of the<br />
coloniesoftenhasdiagnosticsignificance.Incorporatingcomponentsintomediathatalter<br />
the appearance of some colonies <strong>and</strong> not others, depending on the identity of the microorganism,<br />
can enhance this. In this way plate cultures can be used to check that cultures<br />
are pure. Conversely, they can provide visual indication that the microbial population in<br />
the original sample was mixed. Separation of mixed populations into distinct colonies<br />
facilitates subsequent purification by allowing removal of chosen colonies <strong>and</strong> transfer to<br />
fresh medium.<br />
General-purposemediaaredesignedtosupportthegrowthofmanydifferentmicrobial<br />
species. They are useful for detecting heterogeneous populations <strong>and</strong> may be used to<br />
assess total microbial loadings in particular locations. Commonly these are complex<br />
media that contain nutrients that many micro-organisms are able to utilize. Complex<br />
media are not chemically defined. Usually they contain one or more components, which<br />
are general sources of amino acids, proteins, vitamins <strong>and</strong> metal ions. These are<br />
supplemented with specific sources of carbon such as sugars. Examples of general base<br />
ingredients are yeast extract, peptone <strong>and</strong> clarified meat extracts. Some complex media<br />
are tailored to a particular application. For example, in brewing several media<br />
formulations include wort or beer.<br />
Selective media are formulated to promote the growth of certain groups of microorganisms<br />
but not others. They are used in conjunction with selection via control of the<br />
environment, such as manipulation of temperature <strong>and</strong> provision or exclusion of oxygen.<br />
Selective media are used in many ways. Defined media contain only identified <strong>and</strong> often<br />
chemically simple components in combinations that support the growth of some microorganisms<br />
but not others. Media can be made more selective by the incorporation of<br />
inhibitors. These can be very general purpose, for example, the addition of selected<br />
antibioticsthatinhibityeastbut allowthegrowthofbacteria, such asmight beusedinthe<br />
assessment of contamination. Other inhibition protocols might prevent the growth of<br />
many bacteria but allow the growth of yeast. For example, making media acid (
Table 17.5 Some microbiological media used in brewing laboratories (Bridson, 1998)<br />
Medium Application Comments<br />
MYPG General-purpose medium for yeast <strong>and</strong> bacteria. Malt extract, yeast extract, peptone, glucose.<br />
Nutrient agar/broth General-purpose medium for bacteria although many yeasts<br />
will also grow.<br />
WLN (Wallerstein<br />
Laboratory nutrient<br />
medium)<br />
General-purpose medium for yeast but many bacteria will also<br />
grow.<br />
Contains yeast extract, peptone, NaCl <strong>and</strong> Lab-Lemco (clarified<br />
meat extract). Acidification with HCl to pH 4.0 suppresses<br />
growth of many bacteria.<br />
Incorporates pH indicator, bromocresol green which allows<br />
differentiation of some yeast strains from the colour of<br />
colonies.<br />
MYPG ± copper Wild yeast, both Saccharomyces <strong>and</strong> non-Saccharomyces. Addition of hydrated copper sulphate (200 mg.l 1 ) prevents<br />
growth of brewing yeast.<br />
WLD (Wallerstein<br />
Laboratory Differential)<br />
General-purpose medium for bacteria. Addition of cycloheximide (15 mg.l 1 ) suppresses growth of<br />
brewing <strong>and</strong> some wild yeast. Addition of isomerized hop<br />
extract (400 mg.l 1 ) suppresses growth of spore-forming bacilli.<br />
Lysine medium Selective medium for non-Saccharomyces wild yeasts. Synthetic medium in which lysine is the principal source of<br />
nitrogen.<br />
Crystal violet medium Selective medium for some wild yeasts. Crystal violet (20 g/ml 1 ) suppresses the growth of brewing<br />
yeast, variable with others.<br />
Yeast nitrogen base Carbon assimilation medium for yeast. Minimal medium to which various carbon sources are added.<br />
Potato dextrose agar General-purpose medium for yeasts <strong>and</strong> moulds. Contains glucose <strong>and</strong> potato extract.<br />
Hopped wort agar General-purpose medium for yeasts, suppresses growth of<br />
some Gram positive bacteria.<br />
Hopped (1040) wort solidified with agar.<br />
Yeast morphology agar Assessment of yeast colonial morphology. Complex defined medium.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Melibiose medium Differential medium for ale <strong>and</strong> lager strains. Only lager strains grow.<br />
Frateur's medium Differential medium for Acetobacter <strong>and</strong> Gluconobacter spp. Contains yeast extract, ethanol <strong>and</strong> calcium carbonate.<br />
Acetobacter colonies produce clearing due to acid formation<br />
from ethanol. Gluconobacter deposit chalk round clearing due to<br />
continued growth on ethanol with formation of CO2.<br />
Carr's medium Differential medium for Acetobacter <strong>and</strong> Gluconobacter spp. Contains yeast extract, ethanol <strong>and</strong> bromocresol green.<br />
Acetobacter strains produce acid, Gluconobacter strains degrade<br />
acid following prolonged incubation.<br />
Dadd's <strong>and</strong> Martin's<br />
Medium<br />
MRS (Man, de Rogosa,<br />
Sharpe medium)<br />
Medium for isolation of Zymomonas spp. Glucose, yeast extract, peptone supplemented with ethanol (3%<br />
v/v) <strong>and</strong> cycloheximide (50 mgl 1 ). Adjust to pH 4.0 with HCl.<br />
Selective medium for lactic acid bacteria. Buffered peptone, Lab-Lemco. glucose, plus Mg 2+ , Mn 2+ . Made<br />
selective by the addition of 2-phenylethanol <strong>and</strong> cycloheximide.<br />
Modified MRS contains maltose, MRS + beer makes medium<br />
more suitable for application in brewing.<br />
Raka-Ray Recommended selective medium for lactic acid bacteria. Yeast extract, tryptone, liver concentrate, maltose, fructose,<br />
glucose, metal salts. Supplemented with sorbitan mono-oleate,<br />
cycloheximide <strong>and</strong> 2-phenylethanol.<br />
NBB<br />
(Nachweismedium fuÈr<br />
BierschaÈdliche Bacterien)<br />
Selective medium for lactic acid bacteria. Buffered glucose, maltose, peptone, yeast extract, meat extract.<br />
Supplemented with beer <strong>and</strong> an indicator, chlorophenol red.<br />
Pre-reduced PYG agar Detection of Megasphaera. Peptone, yeast extract, glucose medium which is autoclaved <strong>and</strong><br />
placed in anaerobic jar whilst still molten. Allowing agar to set<br />
in anaerobic conditions maintains reducing conditions.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
comparative assessment of results is possible. Successful analysis of the test samples<br />
confirms that brewery laboratories are able to recover, cultivate <strong>and</strong> identify typical<br />
brewery contaminants.<br />
17.4 Microbiological quality assurance<br />
Traditional microbiological analyses do not easily fit into a conventional quality control<br />
system. Analyses based on sampling, plating <strong>and</strong> incubation require three to seven days<br />
to yield results. This is too long. Packaged product must be dispatched to trade as quickly<br />
as possible <strong>and</strong> with a minimum of stock holding. On the other h<strong>and</strong>, the brewery must<br />
ensure that product is wholesome <strong>and</strong> will not deteriorate during its intended shelf-life.<br />
There is an obvious correlation between poor quality <strong>and</strong> lost sales. From a broader<br />
perspective, brewers have an obligation to exercise due diligence <strong>and</strong> guarantee the<br />
harmlessness of their products.<br />
The time lag between sampling <strong>and</strong> result means that product may have reached the<br />
consumer before problems become evident. Fortunately, beers do not support the growth<br />
of pathogens. However, the economic implications of a major product recall could be<br />
catastrophic. To counter this threat there has been a move towards systems quality<br />
assurance. Quality control (QC) is based upon sampling the final product <strong>and</strong> ensuring<br />
that specifications are achieved before dispatch. Quality assurance (QA) ensures that<br />
production systems are sufficiently robust that product quality <strong>and</strong> integrity are<br />
guaranteed. In other words, quality systems are designed to be preventative rather than<br />
based on inspection. Most brewers use a hybrid of QC <strong>and</strong> QA. This is partly a<br />
consequence of the conservatism of the brewing industry. From a microbiological<br />
st<strong>and</strong>point, QA guarantees the integrity of product <strong>and</strong> therefore eliminates the<br />
requirement for final product testing. Most brewers accept this argument but prefer to<br />
follow a more visceral approach <strong>and</strong> dem<strong>and</strong> an apparent clean bill of microbiological<br />
health for finished product.<br />
Microbiological quality assurance systems consist of a number of essential elements,<br />
The process is divided into separate defined sub-processes. A sample plan is constructed<br />
for each sub-process. This considers where samples should be taken, what their size <strong>and</strong><br />
frequency should be <strong>and</strong> what analyses should be performed on them. Results of analyses<br />
must be judged against appropriate specifications. A permanent record of the results must<br />
be maintained. Results are linked with individual batches of product, since there must be<br />
full traceability throughout the entire process from raw materials to finished product.<br />
Complex processes such as brewing require the use of formal quality systems, which<br />
are subject to external <strong>and</strong> independent accreditation. Commonly, the ability to operate to<br />
the ISO 9000 quality st<strong>and</strong>ard is used within the UK brewing industry. Such systems<br />
ensure product wholesomeness <strong>and</strong> also satisfy food safety legislation, for example the<br />
European Food Hygiene Directive <strong>and</strong> United Kingdom Food Safety Act (White, 1994).<br />
Within the quality systems, the formal assessment <strong>and</strong> management of risks is achieved<br />
using HACCP analysis (hazard analysis <strong>and</strong> critical control points). HACCP is integrated<br />
into the broader ISO 9000 system (Kennedy <strong>and</strong> Hargreave, 1997). It is a management<br />
system introduced to ensure food safety (Mundy, 1997). Usually it has been applied<br />
where processes present hazards to health from either physical or chemical agents.<br />
Microbiology has not usually been included except where there might be a health risk<br />
from pathogenic organisms. It is beginning to be used as a method of microbiological<br />
quality assurance where the risk is limited to product wholesomeness.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Hazard analysis<br />
Critical control point (CCP)<br />
analysis<br />
Establish critical limits<br />
Establish monitoring<br />
procedures<br />
Establish corrective<br />
actions<br />
Establish verification<br />
procedures<br />
Establish record keeping <strong>and</strong><br />
documentation procedures<br />
Fig. 17.7 Elements of HACCP analysis.<br />
There are several elements to HACCP analysis. The starting point is to draw up a<br />
detailed flow diagram of the process under consideration. Once constructed, the process<br />
flow diagram must be verified to ensure that all relevant steps have been included. All<br />
HACCP analyses contain seven parts (Fig. 17.7). In the first part each step in the process<br />
is assessed <strong>and</strong> its inherent risks are identified. In the second step the identified risks are<br />
graded to identify those that are critical control points (CCPs). A CCP is a process step<br />
which, if not under proper control, has the potential to cause injury or illness to<br />
consumers. For microbiological control of the brewing process the concept of the CCP is<br />
widened to include a potential to result in the sale of product that is not wholesome. The<br />
third part of the analysis is to set critical limits for each CCP. For example, in the case of<br />
a pasteurized beer, the critical limits would be the time <strong>and</strong> temperatures needed to ensure<br />
that the product is rendered microbiologically stable.<br />
The fourth step is to establish monitoring procedures to ensure that the critical limits<br />
are adhered to <strong>and</strong> the CCP is in control. In the case of the pasteurized product this would<br />
be a permanent record of the times <strong>and</strong> temperatures to which all batches of product had<br />
been exposed. The analysis of this type of step would include procedures to ensure the<br />
veracity of measurements, such as proof of calibration of thermometers. Monitoring<br />
should allow for the identification of trends where the CCP might be moving towards<br />
being out of control. Early identification of such trends allows corrective actions to be<br />
taken.<br />
The fifth part of the analysis establishes corrective actions should a CCP be found to<br />
be out of control. To continue the example of the pasteurized product, this would include<br />
procedures to segregate product, which it was suspected might not have received the<br />
specified heat treatment. The procedures to be followed in these circumstances must be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
detailed in the HACCP plan. They must happen <strong>and</strong> not be subject to discussion. In the<br />
example cited the first priority would be ensure that suspect product could not be sent to<br />
trade. The absolute requirement for reliable systems of traceability <strong>and</strong> labelling can be<br />
readily appreciated. Once suspect product has been isolated a more leisurely examination<br />
of the problem <strong>and</strong> consideration of its fate can be undertaken.<br />
The penultimate step in the HACCP plan is to establish verification procedures. These<br />
are of several types <strong>and</strong> their precise nature depends upon the detail of the process. They<br />
must include two elements. Firstly, a day to day examination of the product to ensure that<br />
it meets pre-established specifications. Secondly, regular <strong>and</strong> preferably independent<br />
audits must be performed to guarantee the integrity of the process. Finally, the entire<br />
HACCP plan must be documented <strong>and</strong> a system of record keeping set up.<br />
17.5 Sampling<br />
An appropriate method must be used for obtaining a sample that is representative of the<br />
part of the process under investigation. The sampling device must operate in a manner<br />
that ensures that there is no contamination from the operator or from any other part of the<br />
process. The sample must be of an appropriate size to ensure that subsequent<br />
microbiological analyses yield significant results. Similarly, the frequency of sampling<br />
must be such that the microbiological quality of the process is assured. The results of<br />
analyses must be compared against predetermined specifications. Any results that fall<br />
outside a specified range must prompt investigations <strong>and</strong> where necessary cause<br />
corrective actions to be made. A record of results must be maintained for the purposes of<br />
traceability <strong>and</strong> for trend analysis.<br />
17.5.1 Sampling devices<br />
Sampling devices must be designed <strong>and</strong> used to ensure that each sample is taken in an<br />
aseptic manner. For permanent installations such as sample cocks that are fitted to vessels or<br />
process pipework, the design <strong>and</strong> location must be such that they are properly cleaned during<br />
CIP. Sample points should be in accessible locations. Prior to withdrawing the sample it<br />
must be possible to sterilize the internal <strong>and</strong> external surfaces of the device. This is best<br />
achieved by steam, but this is rarely available. Portable gas burners can be used to flame<br />
sample cocks, although care must be taken to avoid damage to rubber seals, etc. Where heat<br />
labile components are present external surfaces <strong>and</strong> internal components can be sterilized by<br />
flooding with a 70% v/v solution of methanol or industrial methylated spirits (IMS). IMS<br />
may be ignited (with care!) to facilitate sterilization. All apparatus used to collect samples,<br />
such as tubing <strong>and</strong> containers must be sterile <strong>and</strong> wrapped to prevent contamination.<br />
Sample cocks are of three types. The safest is the diaphragm type. These consist of a<br />
port mounted flush into the surface of the tank or pipe. The port is sealed with a rubber<br />
membrane held in place by a stainless steel nut. The sample is withdrawn by piercing the<br />
membrane with a sterile hypodermic needle attached to a piece of sterile tubing. The<br />
sample is run into a suitable sterile container for transport to the laboratory. These sample<br />
ports must not be heated by flaming <strong>and</strong> the membrane must be sterilized by flooding<br />
with IMS. Membranes must be replaced regularly.<br />
Plug-type sample cocks achieve a seal via a stopper containing a hole that can be<br />
rotated to align with the aperture in the sample port. A combination of good engineering<br />
<strong>and</strong> a thin layer of silicone-type grease achieves a watertight seal between the body of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
device <strong>and</strong> the stopper. Sterilization is achieved by acombination of heat <strong>and</strong> flooding<br />
withIMS.Theoutletfromthesamplingdevicehasaconnectionforahosethroughwhich<br />
the sample is collected. The hose is the most microbiologically suspect part of the<br />
sampling device. Prior to use it must be sterilized by immersion in IMS.<br />
Valve-type sample cocks have ascrew thread mounted spindle that in the closed<br />
position seals the port. Unscrewing the spindle moves it outwards, unseals the port <strong>and</strong><br />
allows liquid to flow. Typically, these sample cocks have two outlets mounted vertically<br />
to each other. Prior to sampling, the whole assembly is sterilized by passing steam<br />
through the top outlet <strong>and</strong> allowing the condensate todrain through the lowerone. Again,<br />
care must be taken to ensure that contamination is not introduced from careless h<strong>and</strong>ling<br />
of hoses. Both plug- <strong>and</strong> valve-type sample cocks are suitable for withdrawing aseptic<br />
liquid samples of any volume. However, in the h<strong>and</strong>s of the unskilled, they are more<br />
prone to result in contamination during sampling compared to diaphragm samplers.<br />
Where comparatively large volumes must be sampled, because the expected microbial<br />
loading is low, it is convenient to use sterile membrane filters. The process liquid must<br />
not contain appreciable amounts of particulate material. Sample volumes are typically 1±<br />
5litres. Usually membranes with apore size of 0.45 Mare used. They are placed within<br />
afilter holder which has appropriate fittings for attachment to the sample point. The<br />
membrane filter is attached, with the usual aseptic precautions, to asample cock (either<br />
diaphragm- or valve-type). Typically ameasured volume is allowed to pass through the<br />
membrane, but slowly, so as to avoid excessive gas breakout. Membranes are collected<br />
still in the holder, returned to the laboratory, removed <strong>and</strong> overlaid onto a plate<br />
containing asuitable solidified medium. Membranes are usually printed with agrid<br />
which, after incubation, facilitates counting of any colonies. By calculation the microbial<br />
count in the original sample is determined. This method can also be used for line drip<br />
tests where asmall proportion of the process flow through apipe is diverted so that it<br />
passes through amembrane, as described. Where very low, or zero counts are expected<br />
membranes may be left in place for several hours.<br />
Themembraneapproachisusedwhereverthereisaneedtoconcentratemicro-organisms<br />
fromalargevolumeofliquidorgas.Suitableapparatusisavailableforasepticprocessingof<br />
liquid samples in the laboratory. This procedure is used for processing pasteurized <strong>and</strong><br />
packaged beers. For example, keg/cask samples <strong>and</strong> whole bottles <strong>and</strong> cans. It can also be<br />
usedforhygienetesting,forexample,confirmationofCIPbytestingoffinalrinseliquors,or<br />
saline rinse samples of steamed kegs. Rinse samples can be analysed using conventional<br />
microbiological techniques. More commonly, the ATP bioluminescence procedure is used<br />
(Section 17.3.1). Alternatively, the cleanliness of surfaces can beassessed by wiping with a<br />
sterile swab. Swabs are rinsed in sterile saline <strong>and</strong> any microbial contamination in the liquid<br />
assessed by conventional techniques or via bioluminesence.<br />
Process gases such as CO2, N2, air or O2 are sampled using bespoke apparatus such as<br />
the Holl<strong>and</strong>aer <strong>and</strong> Dalla Vale device. This consists of a sterilizable housing, which<br />
accommodates a 45 mm plate of nutrient medium. The device is attached aseptically to a<br />
suitable sampling point such that the gas is allowed to flow over the surface of the plate<br />
for a predetermined period of time. Plates are removed <strong>and</strong> incubated to detect growth.<br />
Results are expressed as colony forming units (cfu) per unit time. A criticism of the<br />
method is that prolonged gassing dries out plates <strong>and</strong> leads to underestimates of counts.<br />
This can be avoided by using gassing times of no more than 60 seconds. Alternatively,<br />
the sampling device can be replaced with an Ehrlenmeyer flask sealed with a bung<br />
through which pass inlet <strong>and</strong> outlet tubes. The latter is fitted with a sterile gas filter. The<br />
flask contains sterile saline solution <strong>and</strong> the gas flow to be sampled is allowed to bubble<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
through the liquid for apredetermined time. Microbial contamination is determined from<br />
samples of the saline plated out onto appropriate media.<br />
Environmental microbial surveys should be performed to check loadings in sensitive<br />
areas such as packaging halls or cask racking plants. Atmospheric contamination can be<br />
assessed simply by exposing agar plates of appropriate media in the area of interest. Any<br />
organisms falling onto the plates that grow provide a measure of contamination.<br />
Apparatusdesignedforairsamplingtakestheformofasterilehousinginwhichaplateof<br />
nutrient agar medium is placed. The device, which is portable <strong>and</strong> h<strong>and</strong> held, draws a<br />
known volume of air across the surface of the plate. The plate is removed <strong>and</strong> incubated<br />
<strong>and</strong> resultant microbial colonies counted.<br />
17.6 Disinfection of pitching yeast<br />
Pitching yeast acts as a reservoir for low levels of bacterial contamination. Serial<br />
cropping <strong>and</strong> re-pitching provides a route for infection of successive fermentations.<br />
Providing the level of infection is low <strong>and</strong> the pitching yeast has high viability <strong>and</strong><br />
vitality the effects on fermentation performance <strong>and</strong> beer quality will be small. However,<br />
as described in Section 17.3.3, the common contaminant of pitching yeast, O. proteus,<br />
has the ability to produce potentially carcinogenic nitrosamines from nitrite. To avoid this<br />
hazard it is common <strong>practice</strong> to disinfect pitching yeast prior to pitching.<br />
Several approaches have been used to disinfect yeast each of which rely on agents that<br />
have the ability to selectively kill bacteria but not yeast. Obvious agents for this are<br />
antibiotics, which are used in an analogous manner in selective media. In two early<br />
studies (Gray <strong>and</strong> Kazin, 1946; Case <strong>and</strong> Lyon, 1956) it was demonstrated that tyrothricin<br />
<strong>and</strong> polymyxin B could be used to selectively kill bacteria in the presence of yeast.<br />
Superficially, the use of antibiotics is attractive. No effect on yeast would be expected,<br />
antibiotics are effective at low concentrations <strong>and</strong> their effect would persist into<br />
fermentation. However, this application of antibiotics was suggested at a time when the<br />
risks of selecting for resistant strains of bacteria were not appreciated. It is now<br />
recognized that introduction of antibiotics into foodstuffs is irresponsible <strong>and</strong> this<br />
<strong>practice</strong> has never been implemented. During the 1980s a brief resurgence of this<br />
approach took place. The use of a bacteriocin, nisin, was promoted as a bacterial<br />
disinfection agent for use in brewing (Ogden, 1987). Nisin is a small polypeptide<br />
synthesized by strains of Lactobacillus lactis. It exerts its toxic effects by disrupting the<br />
membranes of susceptible cells. It is used in many food industries, particularly those<br />
producing dairy goods where it is a legally acceptable preservative (E234 in Europe).<br />
Nisin has the same advantages as antibiotics with the additional properties of being<br />
relatively heat stable <strong>and</strong> retaining activity at acid pH. It has the serious disadvantage,<br />
from a brewing st<strong>and</strong>point, in that it is lethal for many Gram positive bacteria but shows<br />
no activity against most Gram negative types. This implies that it would not be suitable<br />
for controlling O. proteus. For this reason, as well as reluctance by brewers to add<br />
`foreign' substances likely to persist into product, the use of nisin has not found favour.<br />
Disinfection of pitching yeast is commonly achieved by acid washing. This process<br />
kills susceptible bacteria, but has no effect on wild yeasts. Yeasts are relatively more<br />
tolerant to acid conditions than most common bacterial contaminants. The process has a<br />
long history, Pasteur suggested that treating with tartaric acid could reduce the levels of<br />
potentially harmful bacteria in pitching yeast. In a typical modern process, yeast slurries<br />
are treated with a food grade acid such that the pH is reduced to approximately pH 2.2<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(+/ 0.1) <strong>and</strong> held for 2±4 h at 3 ëC (+/ 1 ëC). Several acidulants are used, the most<br />
common being phosphoric acid. Mineral acids including sulphuric, nitric <strong>and</strong><br />
hydrochloric are also used. The efficacy of the treatment is improved by incorporating<br />
oxidizing agents such as ammonium persulphate. Thus, ammonium persulphate (0.75%<br />
w/v) plus phosphoric acid, pH 2.8, was more effective than treatment with acid alone at a<br />
pH of 2.2 (Simpson, 1987).<br />
Whether acid washing has deleterious effects on brewing yeast is controversial.<br />
Certainly, the process must be properly controlled. Individual brewers decide on<br />
optimum conditions for their particular yeast strains. Experience suggests that the<br />
conditions described above are close to best <strong>practice</strong>. Addition of the acidulant should be<br />
done in a manner that avoids localized high acid concentration therefore the yeast slurry<br />
should be roused mechanically throughout the whole process <strong>and</strong> the acidulant dosed in<br />
gradually to ensure proper dispersion. Commonly, specific acid washing <strong>and</strong> pitching<br />
tanks are used. These have the advantages that yeast sufficient for pitching is treated with<br />
acid <strong>and</strong> there is no residue to dispose of. Acid washing tanks can be fitted with pH<br />
probes to monitor the addition of acidulant. The siting of the probe in relation to the acid<br />
dosing point should be such that there is no possibility of over-dosing. Where in-tank pH<br />
probes are not used it is preferable to remove a sample of slurry <strong>and</strong> perform a titration to<br />
determine the quantity of acidulant required to achieve the target pH. Pitching into wort<br />
terminates the acid washing via dilution with wort. If there is a process delay such that<br />
pitching is delayed it is best to avoid prolonged exposure to acid conditions by the<br />
addition of food grade sodium hydroxide solution to a pH of 4.0±4.5.<br />
Acid washing has a dramatic effect on the appearance of many pitching yeast slurries.<br />
Commonly there is a marked reduction in viscosity. This has been attributed to<br />
deflocculation. It has been suggested that this might be implicated in a improvement in<br />
fermentation performance by acid washed yeast (Jackson, 1988). Compared with nonacid<br />
washed yeast this author reported that high-gravity worts were fermented more<br />
rapidly, the duration of the diacetyl rest was reduced <strong>and</strong> beers contained lower<br />
concentrations of acetaldehyde. More usually, deterioration of brewing yeast subject to<br />
acid washing has been observed. The effects of washing with acidified ammonium<br />
persulphate on 16 yeast strains were studied (Simpson <strong>and</strong> Hammond, 1989). Although<br />
no gross effects were observed, changes to the cell surface occurred <strong>and</strong> there was<br />
leakage of cellular contents. Using scanning electron microscopy it was demonstrated<br />
that the surface of acid washed yeast acquired characteristic `blebs'. Prolonged treatment<br />
produced slurries that were sticky, possibly indicating some cell lysis.<br />
It was suggested that where yeast was obviously deteriorated it should not be acid<br />
washed. This meant yeast that was heavily contaminated with bacteria, had been<br />
recovered from a previous slow fermentation or from a high-gravity fermentation where<br />
the alcohol concentration exceeded 8% v/v. The reasoning is that stressed yeast is less<br />
well able to withst<strong>and</strong> the additional stress of acid washing. Generally, yeast with a<br />
viability judged by methylene blue staining of less than 80% should not be acid washed.<br />
Where a choice exists, low viability yeast that is obviously stressed will not be chosen for<br />
repitching.<br />
17.7 Cleaning in the brewery<br />
The ability to keep process plant scrupulously clean is an essential prerequisite of good<br />
hygiene. This is a complex <strong>and</strong> frequently neglected part of the brewing process. Too<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
often when new plant <strong>and</strong> processes are installed cleaning is considered as an<br />
afterthought. Tanks <strong>and</strong> complex individual pieces of plant are usually well equipped<br />
with CIP systems. However, cleaning the pipework connecting these vessels is often not<br />
subjected tosufficient scrutiny. There is little benefit in having clean tanks if the attached<br />
pipework is areservoir for taints or infection. Similarly, removable items such as flexible<br />
hoses <strong>and</strong> fittings of various sorts are frequently poorly cleaned <strong>and</strong> act as foci for<br />
infecting otherwise clean plant. For trouble-free operation cleaning must be treated as an<br />
integral part of the brewing process.<br />
Cleaningisrequiredforthreereasons.Firstly,theremovalofanysoilingwhichhasthe<br />
potential to introduce taints into products. Secondly, the removal of soiling which has the<br />
potential to adversely affect the operation of parts of the plant, for example, building up<br />
of scale on heat exchangers. Thirdly, disinfection to eliminate any risk of microbial<br />
spoilage. This section concentrates on the latter aspect of cleaning although inevitably<br />
there is much overlap with the first two. Acomprehensive review of brewery cleaning<br />
regimes <strong>and</strong> some useful definitions can be found in Singh <strong>and</strong> Fisher (1999) (Table<br />
17.6).<br />
The cleaning challenge changes throughout the brewing process (Fig. 17.8). At the<br />
malting stage the spoil is largely restricted to particulate materials <strong>and</strong> plant needs to be<br />
physically clean (although precautions must be taken to control mould growth). During<br />
the wort production stages the soil consists of protein, dextrins, sugars, minerals, tannins<br />
<strong>and</strong> hop materials. Heating during wort boiling can generate scale (beerstone, calcium<br />
oxalate) which becomes baked onto surfaces. There is no microbiological threat <strong>and</strong> so it<br />
is sufficient to render the plant chemically clean. From the wort cooling stage onwards<br />
the soil consists primarily of beer <strong>and</strong> yeast. Throughout all of these stages the process<br />
plant must be both chemically <strong>and</strong> microbiologically clean. This requirement extends into<br />
licensed premises where draught beers are dispensed.<br />
Chemical cleaning relies on treatments capable of removing soils associated with wort<br />
<strong>and</strong> beer. These soils contaminate the surfaces of tanks <strong>and</strong> pipework as films that remain<br />
after product streams have drained away, so with the exception of baked-on scale, most<br />
chemical soils are loosely attached to the surfaces of brewing process plant.<br />
Microbiological soils are more complex <strong>and</strong> potentially much more tenacious. Natural<br />
mixed populations of micro-organisms are commonly found attached to surfaces as<br />
biofilms (Quain, 1999; Stickler, 1999). Indeed, this may be the preferred mode of growth.<br />
Micro-organisms in biofilms, attached by extracellular polysaccharides, are organized so<br />
as to make best use of available nutrients whilst occupying a protected location. Many<br />
surfaces, including stainless steel, are readily colonized. Many common beer spoilage<br />
bacteria are adept at forming biofilms (Czechowski <strong>and</strong> Banner (1992). In addition to the<br />
threat posed in the brewery, biofilms are also important causes of spoilage in licensed<br />
premises. In particular, cleaning of beer dispense lines must be rigorous to ensure that<br />
biofilms are not allowed to form.<br />
Biofilms are of significance to brewing for two reasons. Firstly, once formed they are<br />
difficult to remove, <strong>and</strong> they can act as permanent sources of contamination capable of<br />
seeding successive batches of product. Secondly, <strong>and</strong> perhaps more importantly, analysis<br />
of suspended microbial counts may dramatically underestimate the true magnitude of<br />
contamination. Usually assessment of contamination is based on measurement of the<br />
free-living population. Hygiene tests should be based on the swabbing of surfaces rather<br />
than analysis of rinse water. Cleaning regimes must be sufficiently vigorous to prevent<br />
biofilms forming.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.6 Cleaning regimes ± definitions (adapted from Singh <strong>and</strong> Fisher, 1999)<br />
Term Definition<br />
Physical cleanliness Visually clean.<br />
Chemically clean Cleaned surface imparts no contamination to product. Cleaned surface wets completely with clean water <strong>and</strong> forms a continuous<br />
film.<br />
Microbiologically clean Absence of microbiological contamination.<br />
Cleaning-in-place (CIP) Automatic cleaning of plant without the need to dismantle.<br />
Soil Any substance in the wrong place but typically residues from process liquids <strong>and</strong> solids.<br />
Detergent A chemical cleaning agent, often with surfactant properties for removing soil from surfaces.<br />
Disinfection Treatments that kill micro-oganisms <strong>and</strong> reduce loadings to a desired concentration. It does not imply total killing (sterilization).<br />
Production sterility Term used in industrial microbiology to describe a state in which plant or processes are disinfected to produce conditions which<br />
do not result in spoilage.<br />
Chemical sterilant A chemical disinfectant.<br />
Sanitization Process which is a combination of cleaning <strong>and</strong> disinfection.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Process step Soil St<strong>and</strong>ard of<br />
cleanliness<br />
Malting<br />
Brewhouse<br />
Wort cooling<br />
Fermentation<br />
Conditioning<br />
Bright beer tank<br />
Packaging<br />
Draught dispense<br />
Particulate Physical<br />
Wort<br />
Beer + yeast<br />
Chemical<br />
Chemical +<br />
microbiological<br />
Fig. 17.8 Nature of the soil <strong>and</strong> the st<strong>and</strong>ard of cleanliness required at various stages in the<br />
brewing process.<br />
17.7.1 Range of cleaning operations<br />
Many traditional breweries use open vessels <strong>and</strong> these may still be cleaned manually using<br />
buckets <strong>and</strong> brushes. However, attempts have been made to improve the efficiency of<br />
cleaning, using less manpower. Undoubtedly, this has encouraged the adoption of<br />
automatic cleaning regimes. Thus, traditional open vessels, such as fermenters, are<br />
commonly fitted with detachable hoods that are fitted to enclose the vessel during<br />
automatic cleaning. Detachable fittings such as hose adapters, valves, swing bends,<br />
components of fillers, flexible hoses, etc., may be cleaned manually followed by immersion<br />
in soak baths filled with a suitable disinfectant. Surfaces of complex equipment such as<br />
fillers <strong>and</strong> crowners are commonly cleaned with foams or gels which have excellent wetting<br />
properties <strong>and</strong> an ability to penetrate into hard-to-reach areas. With bottle <strong>and</strong> can filling<br />
machines foaming sprays are often part of an automatic cleaning system of external<br />
surfaces. Environmental cleaning can be a neglected area, particularly with the increased<br />
enclosure of the brewing process. Nevertheless in sensitive areas, such as filling halls <strong>and</strong><br />
yeast rooms, walls, floors <strong>and</strong> ceilings should be cleaned with high-pressure sprayers.<br />
Individual pieces of plant tend to be fitted with dedicated automatic cleaning-in-place<br />
(CIP) systems such as various types of vessel <strong>and</strong> associated equipment such as heat<br />
exchangers, centrifuges <strong>and</strong> filters. Intervening pipework may be cleaned as part of CIP<br />
circuits with vessels or other pieces of plant. Some pipe runs may need to be cleaned<br />
independently. The proximity of cleaning fluids <strong>and</strong> beer streams requires a high st<strong>and</strong>ard<br />
of engineering to ensure that no cross-contamination can occur. In sophisticated <strong>and</strong><br />
complex brewery processes, cleaning <strong>and</strong> other operations are usually fully automatic <strong>and</strong><br />
computer controlled. These systems include fail-safe arrangements to ensure that possible<br />
conflicts are eliminated. On the other h<strong>and</strong> fully automatic systems are often inflexible.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The use of enclosed equipment introduces abarrier between the process <strong>and</strong> the<br />
external environment. Together with improved cleaning this has improved microbiological<br />
st<strong>and</strong>ards. However, these improvements have been costly. CIP systems are<br />
expensive in both capital <strong>and</strong> revenue, therefore it is essential that each system is<br />
appropriatefor itsjob.Onceasystemhas been chosen,itsoperationmust beoptimizedto<br />
maintain abalance between efficiency of cleaning <strong>and</strong> cost of the operation. Several<br />
factors impact on the efficiency of cleaning including the vigour of the delivery of the<br />
cleansing agents to the plant being cleaned. The duration of the treatment, the<br />
temperature <strong>and</strong> nature of the chemical agents used are all influential.<br />
Proper design of brewery plant is an essential prerequisite of good cleaning. Items<br />
such as tanks are fitted with spray-balls, which are designed to distribute jets of cleaning<br />
fluids such that all internal surfaces are cleaned. The jets emanating from spray-balls<br />
must be able to reach all parts of the vessel. Two types of spray-ball are used. Lowpressure<br />
types rely on being placed in an appropriate location to ensure adequate<br />
coverage ofsurfaces.High-pressurespray-ballsare causedtorotatebythe incomingfluid<br />
<strong>and</strong> thereby ensure that all surfaces are covered. Cleaning fluid is collected via the<br />
pipework of the piece of equipment being cleaned <strong>and</strong> re-circulated back through the<br />
spray-balls. Internal metal surfaces must be polished to asuitable st<strong>and</strong>ard to ensure that<br />
therearenocrevicestoharbourmicrobialcells.Similarly,internalwelds<strong>and</strong>fittingssuch<br />
as sample cocks must not provide shadow areas where soil cannot be removed. The rate<br />
of delivery of cleaning fluids must be balanced with rates of draining so that flooding<br />
does not occur <strong>and</strong> thereby negate the scouring action of the spray-balls. Often this is<br />
accomplished by delivering the cleaning agent in aseries of bursts. The duration of<br />
successive cleaning treatments must allow for proper draining to avoid intermixing.<br />
Design <strong>and</strong> the velocity <strong>and</strong> turbulence of cleaning fluids influence cleaning of<br />
pipework. Proper cleaning of pipes requires complete filling with fluid to ensure wetting<br />
of all surfaces. The diameter of the pipe, the velocity of the fluid flow <strong>and</strong> the aspect of<br />
the pipe relative to the fluid flow influence filling. The larger the diameter of the pipe the<br />
higher the flow rate must be to ensure complete filling. As arule of thumb, apipe with a<br />
diameter of 3in. (7.62cm) requires aflow rate of at least 2.2m.sec 1 toensure complete<br />
filling (Fig. 17.9). Vertical runs of pipe work are adequately filled providing the liquid<br />
flow is upwards. Where the fluid flow is vertically downwards even greater velocities are<br />
needed to ensure complete flooding. In the case of athree-inch pipe the vertical downflow<br />
rate must be increased to at least 8.5m.sec 1 toensure complete wetting. This<br />
problem can be ameliorated by slightly reducing the diameter of vertical pipe runs.<br />
Alternatively, non-return valves may be used. Some common problems associated with<br />
pipe runs <strong>and</strong> remedies are shown in Fig. 17.10.<br />
The duration of cleans <strong>and</strong> the temperature at which cleaning agents are delivered is a<br />
balance between efficiency <strong>and</strong> cost. An increase in both parameters is associated with an<br />
increase in cost. The efficacy of alkaline detergents both for removing soil <strong>and</strong> for killing<br />
micro-organisms is increased by elevated temperatures. In order to achieve a given level<br />
of cleanliness <strong>and</strong> microbial kill, there is a rough negative correlation between detergent<br />
strength <strong>and</strong> temperature. Similar relationships exist between both of these parameters<br />
<strong>and</strong> the duration of the clean. Several authors maintain that soil that is formed hot should<br />
be removed hot (Platt, 1986, Singh <strong>and</strong> Fisher, 1999;). On this basis, most of the<br />
brewhouse, wort paraflows <strong>and</strong> pasteurizers should be cleaned hot. However, most areas<br />
which receive heavy yeast soils, such as yeast storage tanks <strong>and</strong> propagation plants, tend<br />
to be cleaned hot. Preliminary rinses <strong>and</strong> acid detergents are usually applied cold,<br />
whereas alkaline detergents are used hot. Irrespective of the operation, the guiding<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Velocity (m/2)<br />
3.5<br />
3.0<br />
2.5<br />
2.0<br />
1.5<br />
1.0<br />
Horizontal pipes full<br />
Horizontal pipes part-full<br />
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Inches<br />
2.54 5.1 7.5 10.2 Centimetres<br />
Fig. 17.9<br />
Pipe diameter<br />
Relationship between pipe diameter <strong>and</strong> velocity of fluid flow required to ensure that<br />
horizontal pipe runs remain full of liquid.<br />
(a)<br />
Inverted pocket<br />
traps air<br />
Joints <strong>and</strong> flanges can<br />
allow air into the system<br />
(b)<br />
High point<br />
traps air<br />
Vertical drop<br />
never fills<br />
Unchoked drain<br />
allows air into system<br />
High-point vent<br />
to bleed air Restricting<br />
vertical drop<br />
helps main to fill<br />
Non-return on drain<br />
keeps air out<br />
Fig. 17.10 Poorly designed pipework (a) <strong>and</strong> some possible improvements (b) (Diagram courtesy<br />
of A. Mielenewski).<br />
principle must always be achievement of a desired clean within an acceptable time <strong>and</strong> at<br />
an acceptable cost. Singh <strong>and</strong> Fisher (1999) conveniently quantified the total cleaning<br />
energy required as the sum of chemical energy (strength of cleaning agent), mechanical<br />
energy <strong>and</strong> thermal energy.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
17.7.2 CIP systems<br />
Several variations are used, all sharing some common features. Cleaning agents are<br />
stored in tanks, which are attached to the pieces of equipment <strong>and</strong> pipework that need to<br />
be cleaned. Cleaning agents are delivered from storage tanks to the plant by dedicated<br />
pumps. Usually asmaller scavenging pump returns fluids to the cleaning tanks to form a<br />
CIP circuit. Additional pipework allows the liquid flow to be directed to drain, as<br />
appropriate. The whole arrangement is termed a CIP set. Very sensitive pieces of<br />
equipment may have their own dedicated CIP sets. More usually, one or asmall number<br />
of CIP sets serve particular parts of the brewery process. The total number of CIP sets<br />
available influences the capacity <strong>and</strong> flexibility of the cleaning operation. Large numbers<br />
of CIP sets increases the cost of the operation. CIP may be manual or automatic. Manual<br />
systems are flexible <strong>and</strong> allow one or asmall number of CIP sets to service many<br />
different parts of the brewery. The nature of the cleaning regime can be varied to suit the<br />
type of soil that needs to be removed. Automatic regimes are inflexible <strong>and</strong> more<br />
expensivethanmanualtypes,howevertheytendtogiveamoreuniformcleaningprocess.<br />
Total dump (total loss or single use) systems do not recover any of the cleaning fluids or<br />
rinses. They are profligate in terms of chemical <strong>and</strong> water usage <strong>and</strong> produce the most<br />
effluent, so they are used only in areas of heavy soil where cleanliness <strong>and</strong> microbiological<br />
hygiene are of paramount importance, e.g., in yeast propagation plant. They may also be<br />
used in conjunction with caustic-based detergents where there is ahigh concentration of<br />
CO 2.Typically, single use CIPsets are dedicated to single pieces of plant.They are located<br />
in close proximity to the piece of equipment they are required to clean.<br />
Partial <strong>and</strong> total recovery systems save some costs by recovering aproportion of the<br />
cleaning fluids. In these systems some or all of the dilute detergent is retained, after<br />
topping up with fresh detergent to ensure that it is of the correct strength it is re-used.<br />
Similarly, aproportion of rinse waters may be returned to the CIP set <strong>and</strong> re-used. These<br />
systems are used where the cleaning task is less onerous, for example, bright beer tanks.<br />
CIP sets have amanual or automatic means of setting the cycle times for individual steps<br />
in the cleaning process. Athermometer <strong>and</strong> thermostat is needed to set the temperature<br />
for hot treatments. Where caustic-based detergents are used it is usual to monitor the<br />
strength by means of aconductivity probe. This may be used as part of an automatic loop<br />
system for dosing concentrated detergent into feed tanks. Care must be taken when<br />
assessing caustic concentration based on conductivity measurements (Section 17.7.5).<br />
CIP treatments consist of a number of individual steps. At the start the plant being<br />
cleaned must be empty so that there is no dilution of cleaning agents. Similarly, the plant<br />
must be fully drained between individual treatments to ensure that there is no intermixing<br />
<strong>and</strong> consequent reduction in the effectiveness of liquids. The process starts with a prerinse<br />
to remove heavy soil. Often pulsed rinsing is use to improve the efficiency of soil<br />
removal. The pre-rinse is sent to drain <strong>and</strong> in recovery systems, the liquid used is often<br />
derived from a recovered final rinse from a previous CIP. The main detergent clean<br />
follows the pre-rinse. This is the longest part of the process in which the liquid is<br />
circulated through the CIP circuit, typically 30±60 minutes. In total dump systems the<br />
detergent is sent to drain. In recovery systems it is retained <strong>and</strong> used in subsequent cleans.<br />
Where alkaline detergents are used an acid rinse may follow to neutralize residual alkali.<br />
The process is completed by a terminal rinse, which typically includes a sterilant.<br />
Terminal sterilants include peracetic acid, chlorine dioxide <strong>and</strong>, now more rarely, sodium<br />
hypochlorite. In some cases it is not appropriate to use a sterilant since it may affect the<br />
process or the beer, for example, propagation plant <strong>and</strong> packaging lines. In these cases<br />
sterility can be assured by a final treatment with wet anaerobic steam.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
17.7.3 Cleaning agents<br />
Soil removal is much improved by the use of detergents. These act synergistically with<br />
physical cleaning processes by loosening soils from surfaces. Several roles performed by<br />
detergents are recognized (Singh <strong>and</strong> Fisher, 1999). Physical effects include wetting of<br />
surfaces,dispersionoflarge agglomeratesofsoiltoformfinelydivided particulatematter<br />
<strong>and</strong> suspension of soil in solution. Chemical effects include dissolution of inorganic<br />
mineral scale by acid detergents, saponification of lipophilic components by alkaline<br />
detergents <strong>and</strong> hydrolysis of proteins.<br />
For the removal of heavy soils the most commonly used detergents are those based on<br />
sodium hydroxide. Typically caustic soda is used at aconcentration of 2±5% w/v at a<br />
temperature of 70±90ëC. It is avery good cleaning agent <strong>and</strong>, when hot, is an effective<br />
biocide <strong>and</strong> efficient at removing biofilms (Czechowski <strong>and</strong> Banner, 1992). It has two<br />
major disadvantages. Caustic soda reacts with CO 2to produce sodium bicarbonate. In<br />
parts of the brewery, such as fermenters where there are high concentrations of CO2, this<br />
can dramatically reduce the effectiveness of caustic detergents. Thus, 1m 3 ofCO2 at 1<br />
bar <strong>and</strong> 20ëC will neutralize 2kg of NaOH (Gingell <strong>and</strong> Bruce, 1998). To avoid this<br />
problem it is necessary either to ensure that vessels <strong>and</strong> pipework are vented to remove<br />
CO2 or to appreciate that losses will occur <strong>and</strong> increase the dosage of caustic detergent<br />
accordingly. Alternatively, acid-based detergents can be used in areas where CO2 will be<br />
aproblem.<br />
Caustic soda reacts with the salts that cause hardness in water to form insoluble<br />
precipitates in the form of calcium carbonate <strong>and</strong> magnesium hydroxide. The precipitates<br />
can accumulate on the surfaces in the form of scale. The problem is exacerbated in areas<br />
subjected to hot cleaning such as parts of the brewhouse. To counter this detergents are<br />
routinely supplemented with chemical additives termed sequestrants. These are of two<br />
types (Table 17.7). Stoichiometric sequestrants form soluble chelates with metal ions.<br />
They break down mineral precipitates in soils or prevent their formation such that scale<br />
formation is minimized. Threshold sequestrants do not prevent the formation of water<br />
hardness precipitates, rather they modify their crystal structure such that they do not<br />
adhere to surfaces <strong>and</strong> form scale.<br />
Acidic detergents have no reaction with CO2 <strong>and</strong>, therefore, are suitable for use with<br />
fermenters <strong>and</strong> associated plant. Commonly they are used cold. Their use for cleaning of<br />
fermenters is not popular since, compared to hot caustic detergents, they are less efficient<br />
at removing heavy soils. However, they do remove scale <strong>and</strong>, where this becomes a<br />
problem, they may be chosen for occasional use. Acid <strong>and</strong> caustic detergents can be used<br />
sequentiallyinheavilysoiledlocationssuchasfermenters.Inthiscase,acausticpre-rinse<br />
isused toremove the worst ofthe soilfollowed by anacid-basedclean. Acidic detergents<br />
are often used for keg washing. The most commonly used acid is phosphoric acid. This is<br />
sometimes supplemented with nitric acid. When used caution must be exercised since<br />
acidic detergents are corrosive.<br />
Theefficacyofbothacid<strong>and</strong>alkalinedetergentscanbeenhancedbytheincorporation<br />
of surfactants. These are molecules that have both hydrophobic (non-polar) <strong>and</strong><br />
hydrophilic (polar) groups. They improve wetting by reducing the surface tension of<br />
liquids. This is achieved by their tendency to adopt aconfiguration in which the nonpolar<br />
groups become aligned towards the surface of liquids <strong>and</strong> the polar groups lie in an<br />
inward-facing orientation. Above certain concentrations (the critical micellar concentration)<br />
the non-polar groups form aggregates termed micelles. These trap lipophilic soil<br />
particles <strong>and</strong> prevent them from re-adhering to surfaces. Water-insoluble surfactants are<br />
defoaming <strong>and</strong> are used where the formation of foam is undesirable, such as bottle<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 17.7 Sequestrants used in CIP cleaning agents (after Singh <strong>and</strong> Fisher, 1999)<br />
Generic sequestrant Examples Action<br />
1. Stoichiometric sequestrants<br />
Amino carboxylic acids Ethylenediaminetetraacetic acid (EDTA)<br />
Nitrilotriacetic acid (NTA)<br />
Chelator of metal ions forming stable complexes <strong>and</strong><br />
helping to remove scale or prevent its formation.<br />
Hydroxycarboxylic acids Derivatives of gluconic acid Chelator of Ca 2+ , Fe 3+ , Al 3+ . Effectiveness is<br />
increased in the presence of free NaOH <strong>and</strong> activity<br />
is maintained at high temperature. They are used in<br />
the brewhouse <strong>and</strong> for bottle washing.<br />
Polyphosphates Sodium tripolyphosphate<br />
Sodium hexametaphosphate<br />
2. Threshold sequestrants<br />
Phosphonic acid derivatives Amino,tris-(methylenephosphonic acid) (ATMP)<br />
1-Hydroxyethane diphosphonic acid (HEDP)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Relatively insoluble chelators often used as<br />
constituents of powdered detergents.<br />
Modifiers of the crystal structure of precipitates so<br />
that they do not adhere to surfaces <strong>and</strong> form scale.<br />
Used in conjunction with caustic soda.
washers. Conversely, water-soluble surfactants produce stable foams <strong>and</strong> are used in<br />
foam cleaners. Surfactants are able to disrupt biological membranes <strong>and</strong> consequently<br />
many are powerful biocides.<br />
Anionic surfactants have negatively charged dissociated groups. They have the<br />
properties of soaps <strong>and</strong> are used as conveyor belt lubricants. Conversely, cationic<br />
surfactants have positively charged dissociated groups. Some are used as corrosion<br />
inhibitors in water treatment systems, for example in pasteurizer sprays. The most common<br />
cationic surfactants are quaternary ammonium compounds. They have disinfectant activity.<br />
Amphoteric surfactants have acharge that is dependent on the pH of the medium they are<br />
added to. They have good biocide activity <strong>and</strong> are used in soak baths or as foam sprays.<br />
Non-ionic surfactants are electrically neutral <strong>and</strong> are used as CIP additives.<br />
Disinfectants (often called sanitizers) have several applications in ensuring microbiological<br />
cleanliness in brewing. Where process sterility is important, roughly from wort<br />
cooling onwards, disinfectants are incorporated into terminal rinses at the end of CIP to<br />
ensure that microbiological loadings remain at low levels. They are used in soak baths for<br />
the same reason. Disinfectants are used in foam spray cleaners <strong>and</strong> are used as additives to<br />
cleaning agents for some manual operations. They are incorporated into non-process water<br />
to prevent microbial growth. Several types are available depending on the application.<br />
Disinfectants should have anumber of attributes (Singh <strong>and</strong> Fisher, 1999). They should<br />
exhibit biocide activity towards abroad spectrum of micro-organisms at low concentration,<br />
be inexpensive <strong>and</strong> have no effect on plant, product or other chemical agents that they may<br />
comeintocontactwith.Nonemeetalloftheseneeds<strong>and</strong>somecompromisesmustbemade.<br />
Several halogen-containing compounds exhibit disinfectant properties due to their<br />
oxidizing properties (see Chapter 3). Those containing chlorine are most commonly used in<br />
brewing. Two forms are used, sodium hypochlorite <strong>and</strong> chlorine dioxide. In alkaline<br />
solutions, sodium hypochlorite is relatively stable <strong>and</strong> it is supplied in this form. At acid pH<br />
(< 5.0) it forms hypochlorous ions, the active form of chlorine which is a powerful biocide. It<br />
is used at a concentration of 50±300 mg.l 1 as a terminal sterilant in rinse liquors. Hypochlorite<br />
has several disadvantages the most serious of which is its ability to cause corrosion. It<br />
is quite unstable especially at low pH (< 5.0) where free chlorine is generated. Sodium<br />
hypochlorite reacts with some organic compounds to form chlorophenols <strong>and</strong> chloramines.<br />
These can produce taints. In addition, trihalomethanes are potential carcinogens.<br />
Some of the disadvantages of hypochlorite are avoided by chlorine dioxide<br />
(Cadwallader, 1992). This is becoming increasingly popular <strong>and</strong> is tending to replace<br />
hypochlorite as the disinfectant of choice for use in terminal rinses. In fact, it is often<br />
used to treat all water used throughout the brewing process (Chapter 3). Usually it is<br />
synthesized in special generating equipment in which HCl is mixed with a solution of<br />
sodium chlorite. The reaction between these two chemicals forms chlorine dioxide, which<br />
is dosed into water at a concentration of 0.1± 0.5 mg.l 1 . At this concentration it is nearly<br />
three times more effective as a biocide than hypochlorite. It is non-toxic, does not<br />
produce taints <strong>and</strong> does not cause corrosion.<br />
Iodine is complexed with surfactants to form preparations termed iodophores. These<br />
are powerful disinfectants <strong>and</strong> are used at concentrations of 10 mg.l 1 available iodine in<br />
soak baths <strong>and</strong> in spray cleaners. At high concentrations they can be corrosive <strong>and</strong> impart<br />
taints <strong>and</strong> so care must be taken to ensure that they are diluted according to the supplied<br />
instructions. Preparations containing bromine (0.5 mg.l 1 available bromine) are used in<br />
the treatment of re-circulating water.<br />
Peracetic acid (CH3 CO OOH) is often incorporated into final rinses as a terminal<br />
sterilant. It is a powerful disinfectant owing to the fact that it decomposes to form acetic<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
acid <strong>and</strong> hydrogen peroxide. It is used cold at a concentration of 75±300 mg.l 1 . It is<br />
difficult to h<strong>and</strong>le in concentrated form, it has an unpleasant odour <strong>and</strong> it is corrosive. For<br />
these reasons the bulk supply must be stored in portable tanks in a bunded area. The<br />
storage area should be vented since over time peracetic acid decomposes <strong>and</strong> generates<br />
gaseous oxygen. The difficulties with h<strong>and</strong>ling mitigate against its use <strong>and</strong> increasingly it<br />
is being replaced with chlorine dioxide.<br />
Non-oxidizing disinfectants used in brewing include quaternary ammonium cationic<br />
surfactants, amphoteric surfactants <strong>and</strong> biguanides. All are used as components of soak<br />
baths. Each has strengths <strong>and</strong> weaknesses. They are too foam active to be used in CIP.<br />
Quaternary ammonium compounds, used at a concentration of around 200 mg.l 1 , show<br />
little activity against Gram negative bacteria <strong>and</strong> they can impart unpleasant fishy taints.<br />
Amphoteric types are equally effective against Gram negative <strong>and</strong> Gram positive bacteria<br />
at concentrations around 1,000 mg.l 1 . They are ineffective against moulds <strong>and</strong> yeast.<br />
Biguanides are derived from guanidine. They are active disinfectants towards a similar<br />
spectrum of micro-organisms as amphoteric surfactants but only within the range of<br />
pH 3.0±9.0. Biguanides are also used to disinfect recirculating water supplies.<br />
17.7.4 Cleaning beer dispense lines<br />
The microbiological integrity of small pack beers must be guaranteed throughout their<br />
shelf-lives. For draught products microbiological quality assurance must extend beyond<br />
the brewery to the premises where they are dispensed. There is no benefit to be gained in<br />
ensuring the highest st<strong>and</strong>ards of quality up to the point at which beer leaves the brewery<br />
without extending this to the management of hygiene within licensed premises. Central to<br />
this is the proper cleaning of dispense lines <strong>and</strong> associated equipment.<br />
Adequate cleaning of beer lines prevents the formation of biofilms. Materials best<br />
suited for use in beer lines have been investigated (Thomas <strong>and</strong> Whitham, 1996). A<br />
variety of materials were investigated for their ability to support biofilms. Nylon <strong>and</strong><br />
medium-density polythene were the least favourable supports <strong>and</strong> these are recommended<br />
for use in beer lines. PVC was the least suitable material. Best <strong>practice</strong> for line<br />
cleaning has been reviewed (Treacher, 1995). Several proprietary line cleaning agents are<br />
available <strong>and</strong> all share common features. They are supplied as concentrates <strong>and</strong> require<br />
dilution with water to a concentration of 1% v/v. They contain sodium hydroxide as the<br />
primary detergent together with sodium hypochlorite (250 mg.l 1 chlorine) for<br />
disinfection purposes <strong>and</strong> a sequestrant for water softening. Several contain indicator<br />
dyes to provide a visual indication that cleaner is present in beer lines.<br />
Beer lines should be cleaned at least every seven days. First the line is flushed with<br />
clean water to eliminate any beer. It is then filled with detergent <strong>and</strong> left to soak for at<br />
least 20 minutes <strong>and</strong> no more than 30 minutes. During this time beer coolers should be<br />
switched off to prevent freezing. During the soak period the detergent action can be<br />
improved by agitation <strong>and</strong> half way through the soak the detergent should be replaced<br />
with a fresh supply. Finally, the lines should be flushed with clean water until no more<br />
cleaner can be detected, e.g., with indicator papers. The water should then be chased<br />
through with beer at which point the line is ready for re-use. During the soaking period it<br />
is essential that the lines <strong>and</strong> dispense equipment are completely filled with cleaning<br />
fluid. Any venting apparatus, etc., should be fully open during the initial filling<br />
procedure. Coupling devices, etc., must be removed <strong>and</strong> cleaned separately before the<br />
lines are soaked. Extending the soak time beyond the recommended period can damage<br />
dispense equipment <strong>and</strong> lead to taints being imparted to beers.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
17.7.5 Validation of CIP<br />
It is essential that procedures are followed to ensure that CIP processes are carried out<br />
correctly. Two types of validation procedure are used. Firstly, checks must be made to<br />
confirm that the conditions employed during the cleaning process were within the<br />
predetermined specification. Secondly, the cleanliness of the plant must be assessed<br />
against predetermined st<strong>and</strong>ards.<br />
CIP checks include cycle times, temperatures <strong>and</strong> the strengths of cleaning chemicals.<br />
Assuring that the correct concentration of caustic soda-based detergent is used is worthy<br />
ofspecialcomment. Commonly,the strengths ofsolutions ofcausticsoda areassessed by<br />
measurement of conductivity. Such readings can be misleading since conversion of<br />
sodium hydroxide to sodium bicarbonate following exposure to carbon dioxide does not<br />
produce a change in this parameter. Preferably, the concentration of caustic soda<br />
solutions should be checked by off-line titration.<br />
The cleanliness of plant after CIP can be checked using ATP bioluminescence<br />
(Section 17.3.1). Where possible this should be supplemented with visual checks to<br />
ensure that spray-balls are functioning correctly <strong>and</strong> no shadow areas exist. Recently the<br />
use of a video camera, termed the `Topscan', mounted in the top of cyclindroconical<br />
fermenters has been recommended for examining vessel cleanliness (Wasmuht <strong>and</strong><br />
Weinzart, 1999). Validation of CIP is essential when new plant is commissioned. Since<br />
CIP is a costly <strong>and</strong> time-consuming process it is necessary to employ conditions that<br />
provide the desired level of cleaning at the lowest cost.<br />
17. 8 References<br />
AMAHA, M., KITABATAKE, K., NAKAGAWA, A., YOSHIDO, J. <strong>and</strong> HARADA, T. (1974) Bull., Brew. Sci., 20, 35.<br />
AMERICAN SOCIETY OF BREWING CHEMISTS (1992) Methods of Analysis, 8th edn, American Soc. Brew.<br />
Chem., St. Paul, Minnesota.<br />
ARCHIBALD, F. S. <strong>and</strong> FRIDOVICH, L. (1981) J. Bacteriol., 146, 928.<br />
BOEIRA, L. S., BRYCE, J. H., STEWART, G. G. <strong>and</strong> FLANNIGAN, B. (1999a) J. Inst. Brew., 105, 366.<br />
BOEIRA, L. S., BRYCE, J. H., STEWART, G. G. <strong>and</strong> FLANNIGAN, B. (1999b) J. Inst. Brew., 105, 376.<br />
BOULTON, C. A. <strong>and</strong> QUAIN, D. E. (2001) <strong>Brewing</strong> Yeast <strong>and</strong> Fermentation, Blackwell Science Ltd.,<br />
Oxford.<br />
BRIDSON, E. Y. (1998) Oxoid Manual, 8th edn, Oxoid Ltd., Hampshire, UK.<br />
CADWALLADER, S. D. (1992) Ferment, 4, 380.<br />
CASE, A. A. <strong>and</strong> LYON, A. I. L. (1956) J. Inst. Brew., 62, 477.<br />
CZECHOWSKI, M. H. <strong>and</strong> BANNER, M. J. (1992) MBAA Tech. Quart., 29, 86.<br />
DE LEY, J., SWINGS, J. <strong>and</strong> GOSSELE, F. (1984) `Key to the genera of the family Acetobacteraceae'. In<br />
Bergey's Manual of Systematic Bacteriology, 9th edn, Vol. 1., N. R. Krieg <strong>and</strong> J. G. Holt, eds,<br />
Williams <strong>and</strong> Wilkins, Baltimore, p. 268.<br />
DIMICHELE, L. J. <strong>and</strong> LLEWIS, M. J. (1993). J. Amer. Soc. Brew. Soc., 51, 63.<br />
ENGELMANN, U. <strong>and</strong> WEISS, N. (1985) Systematics <strong>and</strong> Appl., Microbiol., 6, 287.<br />
EUROPEAN BREWERY CONVENTION (1998), Analytica Microbiologica Vol. II 5th edn, H. C. Verlag,<br />
Germany.<br />
FLANNIGAN, B. (1999). `The microflora of barley <strong>and</strong> malt'. In <strong>Brewing</strong> Microbiology, F. G. Priest <strong>and</strong> I.<br />
Campbell, eds, Aspen Publishers, Inc., Gaithersburg, Maryl<strong>and</strong>, pp. 83±126.<br />
FLEET, G. (1992) Crit. Rev. Biotechnol., 12, 1.<br />
GILLILAND, R. B. (1971) J. Inst. Brew., 77, 276.<br />
GINGELL, K. <strong>and</strong> BRUCE, P. (1998) Proc. 23rd IOB Conv., Asia Pacific Sect., Perth, 134.<br />
GRAY, P. P. <strong>and</strong> KAZIN, A. D. (1946) Wallerstein Lab. Commun., 9, 115.<br />
GUTTERIDGE, C. S. <strong>and</strong> PRIEST, F. G. (1999) `Methods for the rapid identification of microorganisms'. In<br />
<strong>Brewing</strong> Microbiology, F. G. Priest <strong>and</strong> I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg,<br />
Maryl<strong>and</strong>, pp. 239±270.<br />
HAIKARA, A. (1989) Proc. 22nd Cong. Eur. Brew. Conv., Zurich, 537.<br />
HAIKARA, A. <strong>and</strong> LOUNATMAA, K. (1987). Proc. 21st EBC Cong., Madrid, 473.<br />
HAMMOND, J. R. M. <strong>and</strong> ECKERSLEY, B. W (1984) J. Inst. Brew., 90, 167.<br />
HYSERT, D. W., KOVECSES, F. <strong>and</strong> MORRISON, N. M. (1976) J. Am. Soc. Brew. Chem., 34, 145.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
INSTITUTE OF BREWING (1997) Methods of Analysis, Vol. 2, Microbiological, I. O. B., Clarges St.,<br />
London.<br />
JACKSON, A. P. (1988) Tech. Quart. Master Brewers Assoc. Americas, 25, 104.<br />
JOHNSTON, M. A. <strong>and</strong> DELWICHE, E. A. (1965) J. Bacteriol., 90, 347.<br />
KENNEDY, A. L. <strong>and</strong> HARGREAVE, L. (1997) Proc. EBC Symp. Monograph XXVI, Stockholm, 58.<br />
KITABATAKE, K. <strong>and</strong> AMAHA, M. (1974) Bull. Brew. Sci., 20, 1.<br />
MAGLIANI, W., CONTI, S., GERLONI, M., BERTOLOTTI, D. <strong>and</strong> POLONELLI, L. (1997) Clin. Microbiol. Rev.,<br />
10, 369.<br />
MAGNUS, C. A., INGLEDEW, W. M. <strong>and</strong> CASEY, G. P. (1986) J. Am. Soc. Brew. Chem., 44, 158.<br />
MAULE, A. P. <strong>and</strong> THOMAS, P. D. (1973) J. Inst. Brew., 79, 137.<br />
MCCAIG, R. <strong>and</strong> MORRISON, M. (1984) J. Am. Soc. Brew. Soc., 42, 23.<br />
MUNDY, A. P. (1997) Proc. EBC Symp. Monograph XXVI, Stockholm, 141.<br />
OGDEN, K (1987) J. Inst. Brew., 93, 302.<br />
OGDEN, K. (1993). J. Inst. Brew., 99, 389.<br />
PHILLISKIRK, G. <strong>and</strong> YOUNG, T. W. (1975) Antonie Van Leeuwenhoek, 41, 147.<br />
PLATT, D. (1986) Brew. Dist. Internat., 16, 20.<br />
PRESCOTT, L. M., HARLEY, J. P. <strong>and</strong> KLEIN, D. A. (1996) Microbiology, McGraw-Hill, USA.<br />
PRIEST, F. G. (1999) `Gram positive brewery bacteria'. In <strong>Brewing</strong> Microbiology, F. G. Priest <strong>and</strong> I.<br />
Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryl<strong>and</strong>, pp. 127±161.<br />
QUAIN, D. E. (1999) Proc. 27th EBC Cong., Cannes, 239.<br />
RUSSELL, I. <strong>and</strong> DOWHANICK, T. M. (1999). `Rapid detection of microbial spoilage'. In <strong>Brewing</strong><br />
Microbiology, F. G. Priest <strong>and</strong> I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryl<strong>and</strong>,<br />
pp. 209±235.<br />
RYDER, D. S., MURRAY, J. P. <strong>and</strong> STEWART, M. (1978) MBAA Tech. Quart., 15, 79.<br />
SAMI, M., YAMASHITA, H. <strong>and</strong> HORONO T. (1997a) J. Ferment. Bioengin., 84, 1.<br />
SAMI, M. YAMASHITA, H., KADOKURA, H., KITAMOTO, K., YODA, K. <strong>and</strong> YAMASAKI, M. (1997b) J. Am. Soc.<br />
Brew. Chem., 55, 137.<br />
SHIMWELL, J. L. (1936) J. Inst. Brew. 42, 585.<br />
SIMPSON, W. J. (1987) J. Inst. Brew., 93, 313.<br />
SIMPSON, W. J. (1993) J. Inst. Brew., 99, 405.<br />
SIMPSON, W. J. (1999) Brewers' Guardian, May, 24.<br />
SIMPSON, W. J. <strong>and</strong> FERNANDEZ, J. L. (1992) Lett. Appl. Microbiol., 14, 13.<br />
SIMPSON W. J. <strong>and</strong> FERNANDEZ, J. L. (1994) J. Am. Soc. Brew. Chem., 52, 9.<br />
SIMPSON, W. J. <strong>and</strong> HAMMOND, J. R. M. (1989) J. Inst. Brew., 95, 347.<br />
SIMSON, W. J. <strong>and</strong> HAMMOND, J. R. M. (1991) Proc. 23rd EBC Cong., Lisbon, 185.<br />
SIMPSON, W. J., HAMMOND, J. R. M., THURSTON, P. A. <strong>and</strong> KYRIAKIDES, A. L. (1989) Proc. 23rd EBC Cong.,<br />
Lisbon, 185.<br />
SINGH, M. <strong>and</strong> FISHER, J. (1999). `Cleaning <strong>and</strong> disinfection in the brewing industry'. In <strong>Brewing</strong><br />
Microbiology, F. G. Priest <strong>and</strong> I. Campbell, ed., Aspen Publishers, Inc., Gaithersburg, Maryl<strong>and</strong>,<br />
pp. 271±300.<br />
SMITH, N. A. (1994) J. Inst. Brew., 100, 347.<br />
SPECKMAN, R. A. <strong>and</strong> COLLINS, E. B. (1973) Appl. Microbiol., 26, 744.<br />
STICKLER, D. (1999) Curr. Opinion in Microbiol., 2, 270.<br />
TAKAHASHI, T., NAKAKITA, Y., MONJI, Y., WATARI, J. <strong>and</strong> SHINOSTUKA, K. (1999a) Proc. 27th Cong. EBC,<br />
Cannes, 259.<br />
TAKAHASHI, T., NAKAKITA, Y., SUGIYAMA, H., SHIGYO, T. <strong>and</strong> SHINOTSUKA, K. (1999b) J. Biosci., Bioeng.,<br />
88, 500.<br />
THOMAS, K. <strong>and</strong> WHITHAM, H. (1996) Proc. EBC Symposium, Edinburgh, Monograph XXV, 124.<br />
TREACHER, K. (1995) Brew. Guard., Aug., 19.<br />
TSUCHIYA, Y., KANO, Y. <strong>and</strong> KOSHINO, S. (1992a) J. Am. Soc. Brew. Chem., 51, 40.<br />
TSUCHIYA, Y., KANEDA, H., KANO, Y. <strong>and</strong> KOSHINO, S. (1992b) J. Am. Soc. Brew. Chem., 51, 64.<br />
VAN VUUREN, H. J. J. (1999) `Gram negative spoilage bacteria'. In <strong>Brewing</strong> Microbiology, F. G. Priest <strong>and</strong><br />
I. Campbell, eds., Aspen Publishers, Inc., Gaithersburg, Maryl<strong>and</strong>, pp. 163±191.<br />
VERZELE, M. (1986) J. Inst. Brew., 92, 32.<br />
WAINWRIGHT, T. (1972) Brewers' Digest, 47, 78.<br />
WASMUHT, K. <strong>and</strong> WEINZART, M. (1999) Brauwelt Internat., 17, 512.<br />
WHITE, F. H. (1994) Proc. EBC Symp., Monograph XXI, Nutfield, 2.<br />
YASUI, T. <strong>and</strong> YODA, K. (1997) Appl. Environ. Microbiol., 63, 4528.<br />
YOUNG, T. W. (1987) `Killer yeasts'. In The Yeasts, 2nd edn, Vol. 2. A. H. Rose <strong>and</strong> J. S. Harrison, eds,<br />
Academic Press, London, pp. 131±164.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
18<br />
Brewhouses: types, control <strong>and</strong> economy<br />
18.1 Introduction<br />
The objective of this chapter is to consider the evolution of brewhouses <strong>and</strong> to consider<br />
briefly the diversity of design of breweries that are operated today. Control systems <strong>and</strong><br />
economic aspects of operation will be considered.<br />
Brewhouse equipment must operate in a reproducible way to yield wort with a<br />
composition that is expected from an underst<strong>and</strong>ing of the biochemistry of the process<br />
(Chapter 4). In large breweries (say, greater than 500,000hl, or 300,000 imp. brl annual<br />
volume) of great importance is the ease with which brewhouse equipment can be<br />
automated, however, there is a warning here. Biochemists study enzymes in model<br />
systems at controlled pH value <strong>and</strong> temperature often approaching ideal conditions. The<br />
environment in abrewery mash is not ideal <strong>and</strong> sometimes unpredictable results will<br />
occur. The level of automation should, therefore, be such that manual intervention can<br />
easily take place when things go wrong. This will avoid the production of a series of<br />
worts of poor quality with consequent adverse quality <strong>and</strong> economic effects.<br />
A good brewhouse, then, is a set of vessels which will allow the biochemical reactions<br />
of brewing to take place in a controlled way as close as possible to the ideal environment<br />
in which the enzymes have been mostly studied. Engineers, therefore, should utilize the<br />
experience of brewers <strong>and</strong> scientists so that the equipment they design <strong>and</strong> sell will<br />
effectively support the critical enzyme activities needed for wort production.<br />
18.2 History of brewhouse development<br />
Visiting different breweries reveals considerable variation in the methods of construction<br />
<strong>and</strong> the layout of the plant. Pioneering work in the late 19th century by Horace Brown<br />
<strong>and</strong> others (Anderson, 1993) stimulated brewers to think even more carefully about<br />
brewhouse design. Economic pressures also became more acute at this time <strong>and</strong><br />
competition in the brewing industry became a reality. Directors of Companies dem<strong>and</strong>ed<br />
adequate financial returns on the capital employed in the brewery <strong>and</strong> looked to make<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
profit from the brewing of beer. In the late 19th century most brewers were brewers for<br />
sale <strong>and</strong> did not own the premises in which beer was sold. The cost of beer production,<br />
therefore, now came to be understood <strong>and</strong> its control was vital to Company success. This<br />
happened in Europe, particularly in the UK <strong>and</strong> in North America (Anderson, 1993).<br />
Brewers were thus forced to look critically at the design of the plant they were using to<br />
produce their beer.<br />
To lower costs it was appreciated that economies of scale were important <strong>and</strong> larger<br />
breweries (>300,000hl pa) were built, mainly in the UK. In the middle of the 19th<br />
century in Germany, however, there were around 15,000 breweries <strong>and</strong> many of these<br />
were extremely small, concentrating on brewing for one retail outlet. Some of these<br />
breweries, only partially modernized, exist today. However, it has been much easier to<br />
introduce new technology <strong>and</strong> lower costs in larger breweries <strong>and</strong> this has been the<br />
fashion in the UK, North America, Sc<strong>and</strong>inavia <strong>and</strong> Japan throughout the 20th century.<br />
Developments in the USA were, of course, severely restricted by the introduction of the<br />
prohibition of alcohol that lasted from 1919 to 1933.<br />
The choice of raw materials for brewing has also influenced developments. Brewers<br />
established themselves commercially in the German States in the 15th century <strong>and</strong><br />
municipal laws closely controlled their activities. To avoid the use of substitute<br />
materials,someofwhichhadbeeninjurioustohealth(someflavourings replacinghops),<br />
the purity law, the Reinheitsgebot, was introduced in 1516 <strong>and</strong> has applied throughout<br />
Germany since 1906. The law declares that beer can be made only from barley, water<br />
<strong>and</strong> hops. This has strongly influenced the development of brewing technology in<br />
Germanyinthelast100years.TheReinheitsgebotappliedinmanycountriesintheearly<br />
stages of brewing development, particularly as so many German brewers emigrated. But<br />
by 1870 American brewers had experimented with maize <strong>and</strong> rice mash tun adjuncts to<br />
economic advantage <strong>and</strong> did not go back to barley malt alone. In the UK there was a<br />
plentiful supply of sugar from the West Indies <strong>and</strong> again this was used to advantage to<br />
lower costs. Against this background the different trends in brewhouse improvement<br />
must be set.<br />
18.2.1 The tower brewery lay-out<br />
The development of the steam engine by James Watt in 1765 provided the basis for the<br />
industrialrevolutioninVictorianBritain.Brewerswerequicktorelatethescientificwork<br />
ofHoraceBrowntoWatt'sgreatinvention<strong>and</strong>alsotothepioneering workofJouleinthe<br />
early 1840s (working in the laboratory of the family brewery in Salford) leading to the<br />
First Law of Thermodynamics, <strong>and</strong> hence the concept of the mechanical equivalence of<br />
heat.<br />
Brewers in the 19th century quickly realized that making more use of gravity could<br />
lower costs in the brewhouse. The concept of the `Tower' brewery was born with<br />
consequent savings in energy <strong>and</strong> manual labour. The need for excessive pumping of<br />
worts was eliminated, resulting in quality improvements as potential sources of infection<br />
<strong>and</strong> oxidation were lessened.<br />
Tower breweries had aflat roof supporting water cisterns <strong>and</strong> tanks (Fig. 18.1). A<br />
considerable weight had to be supported on a small area. The buildings, therefore, had to<br />
be very substantially constructed with solid foundations <strong>and</strong> thick walls. These methods<br />
of construction were immediately available in the mid to late 19th century through to the<br />
1950s (Jeffery, 1956), when cheaper building materials were developed. This meant that<br />
the excessive dem<strong>and</strong>s of the civil construction of the Tower brewery became very<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Mill<br />
Grist<br />
case<br />
A<br />
B<br />
C<br />
Dust<br />
quencher<br />
Hot liquor<br />
Malt screen<br />
Mash tun<br />
Slotted<br />
false bottom<br />
Malt store<br />
Cold liquor tank<br />
tanks<br />
A – Mashing liquor<br />
B – Sparge liquor<br />
C – Underlet liquor<br />
Steel’s masher<br />
Sparge arms<br />
Rakes<br />
Spend<br />
safe<br />
Underback<br />
Hop store<br />
Cellar<br />
Malt<br />
hopper<br />
Condenser<br />
Steam<br />
inlet<br />
Wort<br />
pump<br />
Primings<br />
dissolving<br />
vessel<br />
Copper<br />
Wort<br />
pump<br />
Steam<br />
jacket<br />
Sugar dissolving<br />
vessel<br />
Heater<br />
Steam outlet<br />
Hop sparge<br />
Hop back<br />
Slotted<br />
false bottom<br />
Wort<br />
receiver<br />
Primings cooler<br />
Primings storage<br />
vessel<br />
Enclosed<br />
wort<br />
refrigerator<br />
Spreader<br />
W.W. attemperator<br />
liquor: inlets <strong>and</strong> outlets<br />
Suction yeast Parachute<br />
skimmer yeast skimmer<br />
W W<br />
Attemperator<br />
Fermenting<br />
Racking<br />
back<br />
vessels<br />
Yeast<br />
collecting<br />
vessels<br />
Yeast<br />
press<br />
Fig. 18.1 Sectional representation of a traditional small ale `Tower' brewery, operating with one<br />
elevation of the wort to the Wort Boiler (Jeffrey, 1956).<br />
expensive <strong>and</strong> few were built after the 1960s. In most Tower breweries pumping only<br />
once elevated the wort. This was usually after separation from the extracted malt when<br />
the wort would be pumped to the wort boiler. The wort boiler would be situated at the top<br />
of the brewery <strong>and</strong> then the wort would fall by gravity to the hop separator (hop back) <strong>and</strong><br />
then to a wort receiver prior to cooling.<br />
These Tower brewhouses were built to be as fire resistant as possible <strong>and</strong> the use of<br />
wood was avoided. The buildings were essentially brick <strong>and</strong> concrete with steel being<br />
used in later developments. One of the essentials of a Tower brewery was a hoist or lift<br />
from the bottom to the top of the building. This in itself created a problem for fire<br />
prevention <strong>and</strong> the lift had to be installed in a brick shaft with self-closing doors of sheet<br />
iron on each floor.<br />
Tower breweries also had very small vessels by the st<strong>and</strong>ards of the late 20th <strong>and</strong> early<br />
21st centuries. A large wort boiler would be no more than 325 hl (200 imp. brl) in<br />
capacity whereas today a wort boiler could have a volume of 1,000 hl (600 imp. brl) or<br />
more. In big brewing companies economies of scale tend to dem<strong>and</strong> large vessels <strong>and</strong><br />
these breweries are often producing huge volumes of one or two br<strong>and</strong>s. A Tower<br />
brewery would simply be too expensive to construct to satisfy these requirements. For a<br />
modern brewery a horizontal lay-out using cheaper building materials offers a lower-cost<br />
solution.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
18.2.2 The horizontal brewery lay-out<br />
As often is the case when examining developments in an historical context no system is<br />
found to be entirely designed to one set of rules. We therefore have breweries that have<br />
partial characteristics of vertical or horizontal design. Small breweries, particularly on<br />
mainl<strong>and</strong> Europe, were often of two-vessel design (Fig. 18.2) having one vessel as a mash<br />
tun <strong>and</strong> a separator <strong>and</strong> a second vessel as a mash cooker <strong>and</strong> wort kettle. The lay-out of<br />
these breweries was usually horizontal. It was natural, therefore, that when these plants<br />
were extended to four or more vessels that these were laid out on the horizontal. The<br />
construction of the building to house the vessels is made much simpler in this situation.<br />
This usually means a steel-framed building with brickwork to first-floor level <strong>and</strong> then<br />
some form of mild steel profile cladding <strong>and</strong> glass to the upper levels. Fireproofing is<br />
simpler <strong>and</strong> there is no need for a complex design of lift shaft. Costs of construction are<br />
much reduced <strong>and</strong> the money can be spent on ensuring the best quality of vessels,<br />
pipework <strong>and</strong> control systems. The Tower brewery allowed savings in manpower <strong>and</strong><br />
energy use but this was overtaken by high building costs <strong>and</strong> the subsequent ease of<br />
automation of very large horizontally laid out plant allowed further savings, which more<br />
than compensated for any energy increases.<br />
Mash<br />
mixer<br />
Grist<br />
Mash or<br />
wort boiler<br />
Lauter tun<br />
Section<br />
Fig. 18.2 Two-vessel brewhouse with mash mixing vessel on the left, mash <strong>and</strong> wort boiler in the<br />
centre <strong>and</strong> lauter tun on the right (Briggs et al., 1981).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Plan
18.3 Types of modern brewhouses<br />
Amodern horizontal brewery will have acombination of vessels:<br />
· amash conversion vessel or tun<br />
· amash cooker<br />
· awort separation device such as alauter tun or mash filter<br />
· awort boiler<br />
To these vessels could be added:<br />
· aheated wort buffer tank<br />
· ahop separator such as awhirlpool.<br />
ThedetailedoperationofallthesevesselsisconsideredinChapter6<strong>and</strong>therelated<strong>science</strong><br />
is discussedinChapter 4.Fromthese Chapters theessential technology ofthelargemodern<br />
brewhouse emerges <strong>and</strong> this is obviously the most significant for brewing: <strong>science</strong> <strong>and</strong><br />
<strong>practice</strong>. Different types of brewhouse are, however, also successful in the 21st century.<br />
Brewhouses have to produce a volume of wort of the right quality to support<br />
fermentation as frequently as is required by the market in which the brewing company is<br />
operating. This, of course, derives from the br<strong>and</strong>s of beer the brewery is expected to<br />
produce. The requirements from amicro- or pub brewery are very different from a<br />
brewery producing very large volumes of anational or international br<strong>and</strong>. Abrewery<br />
producing high volume br<strong>and</strong>s might be of asize to produce between 3.5 <strong>and</strong> 10 million<br />
hl (2 to 6million imp. brl) of beer per year. There then would be arange down in size<br />
throughmedium-sizedplantsof0.5to0.8millionhl(0.3to0.5millionimp.brl)tomicrobreweries<br />
as small as 1,000hl (600 imp. brl) per year or even less.<br />
All these types of plant are important in the brewing industry of today; this is from<br />
where the great variety of beer that we enjoy derives <strong>and</strong> this has been an important change<br />
in the last 20 years. We have seen the emergence of the global br<strong>and</strong> produced in many<br />
countries to very strictly defined principles <strong>and</strong> the rise of the so-called boutique or niche<br />
br<strong>and</strong>s produced to satisfy avery local dem<strong>and</strong>. There have been casualties on the way <strong>and</strong><br />
many breweries in Europe <strong>and</strong> North America <strong>and</strong> other parts of the developed world have<br />
closed. There is almost apolarization from the global to the local which large companies<br />
try tobridgewiththe adage`thinklocal actglobal' butitis oftendifficult toreconcile these<br />
conflicting factors. This has implications for astudy of the diversity of brewhouses in the<br />
21st century. The driving force for any business, however, is that it must be financially<br />
successful <strong>and</strong> the yield <strong>and</strong> quality of wort from abrewhouse is important to both the<br />
brewer of the global br<strong>and</strong> <strong>and</strong> the micro-brewer serving avery local community.<br />
In large brewhouses (Chapter 6), brewers may be trying to cope with the dem<strong>and</strong>s of up<br />
to 14 brews in one day. There are very different dem<strong>and</strong>s facing brewers operating other<br />
brewhouses. These other brewhouses can be broadly categorized as experimental<br />
brewhouses <strong>and</strong> micro- or pub breweries. Beer produced in the former category does not<br />
alwayshavetobecommerciallyacceptablewhereasbeerfromapubbrewerymusthavethe<br />
highest acceptability. The requirements of the two brewhouses are therefore quite different.<br />
18.3.1 Experimental brewhouses<br />
These brewhouses can be used solely for training purposes or can be used for a<br />
combination of training <strong>and</strong> raw material evaluation <strong>and</strong> new product development. It is<br />
often difficult to flavour match beers produced on the small scale to those produced on<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the production plant. New product development, therefore, is frequently carried out on<br />
the large scale <strong>and</strong> is correspondingly very expensive! However, raw material evaluations<br />
such as those of malt prepared from a new barley variety or new hop varieties are usually<br />
successful. Indeed in the UK, pilot scale trials coordinated by The Institute <strong>and</strong> Guild of<br />
<strong>Brewing</strong> are crucial in gaining approval of new varieties of both barley <strong>and</strong> hops. These<br />
trials have operated for the last 30 years <strong>and</strong> have prevented barleys of poor brewing<br />
quality becoming widely grown. Beers produced in experimental trials can sometimes be<br />
blended back into mainstream brews at say 10% to minimize costs.<br />
Experimental breweries can vary considerably in size. Breweries from 5 to 200 litres<br />
have been described (De Clerk <strong>and</strong> De Clerk, 1965; Baetsle, 1983). And there is also the<br />
genuine pilot brewery that can be between 10 <strong>and</strong> 100 hl in brewlength (Moll <strong>and</strong><br />
Midoux, 1985). The greater the size means the greater the likelihood results of trial brews<br />
will be closer to commercial results. Most international brewers will have pilot plant<br />
facilities <strong>and</strong> all training institutes <strong>and</strong> research organizations will have this type of<br />
equipment. Results must be interpreted with care <strong>and</strong> the skill <strong>and</strong> experience of the<br />
brewer in charge of the equipment is of major importance. The brewing equipment must<br />
be capable of reproducible operation <strong>and</strong> it is highly desirable to carry out a statistical<br />
evaluation so that the value of the least significant difference between the various<br />
parameters is known. Only in this way will truly meaningful results be obtained.<br />
18.3.2 Micro- <strong>and</strong> pub breweries<br />
The international br<strong>and</strong>s of beer are highly specified <strong>and</strong> extremely consistent. To ensure<br />
this consistency <strong>and</strong> to provide the reproducibility necessary to brew the beers all over the<br />
world these beers tend not to have strong flavours. They tend to be bl<strong>and</strong> <strong>and</strong> in particular<br />
have low bitterness <strong>and</strong> lack hop character. The international br<strong>and</strong>s are also tending to<br />
taste more <strong>and</strong> more similar. Beer drinkers now dem<strong>and</strong> more choice in their beers <strong>and</strong><br />
have looked for more original flavours, usually with higher levels of bitterness. This has<br />
been particularly driven by changes in the USA (Lewis <strong>and</strong> Lewis, 1996) since 1985 when<br />
the micro-breweries started to emerge. This was partly as a result of legislation in 1982 that<br />
legalized `Brewpubs'. Micro-brewing in the USA still occupies a small volume but it has<br />
grown at 50% per year since 1985 to take over 2% of the whole market. The structure of the<br />
industry is complex with a large variation in the size of individual companies but genuine<br />
micro-brewers are driven by real passion for what they do. Consequently their enthusiasm<br />
has had an impact on the major brewers who have sought to cash in on market opportunities<br />
by introducing their own boutique br<strong>and</strong>s often under disguised labelling. This has further<br />
fuelled competition <strong>and</strong> development. The strongest markets for micro-brewers in the USA<br />
are in California, Oregon, New Engl<strong>and</strong> <strong>and</strong> Florida.<br />
Most micro-breweries in the USA use infusion mash tuns for extraction of the malt<br />
<strong>and</strong> wort separation. This equipment is relatively low cost <strong>and</strong> allows a link to the British<br />
tradition of ale brewing. Indeed this is the normal choice throughout the world.<br />
Frequently the brewing equipment is on display to the public consuming the beer. The<br />
breweries vary considerably in size often from about 10 to 50 hl (6 to 30 imp. brl) in<br />
brewlength with three to five brews per week. Consumers will tolerate some variation in<br />
the beer in the interests of presumed authenticity but will not tolerate any hint of<br />
infection. Sanitation <strong>and</strong> microbiological control in these breweries is therefore of<br />
paramount importance <strong>and</strong> if this fails then this is the most frequent cause of<br />
unacceptable beer. Modern micro-brewing apparatus usually has associated CIP<br />
equipment <strong>and</strong> this must be used rigorously. Yeast must also be checked before pitching<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 18.1 Comparison of analyses of generic national lager <strong>and</strong> generic micro-brewed lager in<br />
the USA (Lewis <strong>and</strong> Lewis, 1996)<br />
Parameter National lager Micro-brewed lager<br />
Original gravity (ëP) 11.5 15.5<br />
Alcohol by weight (%) 3.8 6.0<br />
Bitterness (IBU) 12 35<br />
Colour (ëEBC) 4 25<br />
Table 18.2 Japanese consumption <strong>and</strong> market share of beer <strong>and</strong> happoshu (Malone, 2001)<br />
Volume (000 hl) 1996 1997 1998 1999 2000<br />
Beer & happoshu 72,485 72,095 72,351 72,264 71,764<br />
Beer 69,770 67,929 62,561 58,327 55,510<br />
Happoshu 2,715 4,167 9,790 13,937 16,254<br />
Market share (%) 1996 1997 1998 1999 2000<br />
Beer 96 94 87 81 77<br />
Happoshu 4 6 14 19 23<br />
for purity. Bottling equipment is often of poor quality. In addition to ensuring that it is<br />
effectively cleaned, every effort must be made to limit the occurrence of headspace air<br />
that has exceeded 4 ml in some cases with disastrous consequences causing oxidation <strong>and</strong><br />
rapid flavour deterioration. Micro-brewed beers are often prepared with 100% malt <strong>and</strong><br />
as such may be inherently less stable than national beers. This is sometimes offset by<br />
their normally much higher hop content (Table 18.1).<br />
There have been some interesting developments in Japan (Malone, 2001), where a<br />
mature beer market was in need of some innovation. The licensing laws in Japan were<br />
changed in 1994 allowing breweries to make a minimum of 600 hl (360 imp. brl) of beer<br />
per year. There are now around 300 small breweries producing local beer (ji-biru). To<br />
some extent growth of these breweries has been stimulated by the growth of local sake<br />
breweries (ji-zake). There has also been very significant growth in `happoshu' (sparkling<br />
drinks). These products have less than 25% malt in the grist <strong>and</strong> are subjected to much<br />
lower rates of duty. The drinks thus retail at lower prices <strong>and</strong> have grown in popularity<br />
(Table 18.2). This has created a market opportunity for more flavoursome micro-brewed<br />
beers that will grow, particularly if the Government moves to reduce the differences in<br />
duty payments between happoshu <strong>and</strong> beer.<br />
The majority of the small breweries in Japan produce top-fermented beers but some<br />
have tried to produce German-style lagers in this way <strong>and</strong> have found this difficult, with<br />
dire commercial consequences. A shakeout of breweries is now under way but it seems<br />
likely that growth above the current 0.5% market share of ji-biru will continue.<br />
Throughout the world the most successful micro-breweries tend to be those producing<br />
draught products for consumption with a very short shelf-life of no more than four weeks.<br />
Problems arise when producing small pack products <strong>and</strong> trying to extend the shelf-life to<br />
six months <strong>and</strong> beyond. This usually results in oxidized beers that fail dismally when<br />
compared to beers brewed in large breweries. This comparison is best avoided. Micro<strong>and</strong><br />
pub-brewed beers taste best when drunk fresh <strong>and</strong> this is the unique selling point to<br />
exploit. The most successful of these operations keep things simple <strong>and</strong> use a two-vessel<br />
brewery with a mash tun <strong>and</strong> a kettle that can double as a mash kettle <strong>and</strong> wort kettle as<br />
necessary.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
As the 21st century progresses it will be interesting to see how the micro-brewing<br />
industry develops. International br<strong>and</strong>s of beer will continue to be heavily advertised<br />
throughout the world. To compete, local beers will need to be brewed in scrupulously<br />
clean plants by dedicated, well-trained professionals, they will have strong <strong>and</strong><br />
characteristic flavours <strong>and</strong> be drunk as fresh as is possible.<br />
18.4 Control of brewhouse operations<br />
A key feature of any good brewhouse is reproducibility. An environment must be<br />
provided in which the biochemical reactions can take place in a controlled <strong>and</strong> optimized<br />
way. This was originally achieved by close operator involvement in constant checking of<br />
times <strong>and</strong> temperatures <strong>and</strong> taking remedial action where appropriate. Things are now<br />
different. There are very few operators, maybe only two on a 12-hour shift scheduled for<br />
12 brews a day, <strong>and</strong> they are likely to be led by a team leader controlling the whole of the<br />
process from raw material intake to the end of fermentation. The departmental manager<br />
may be a remote figure frequently most concerned with achieving the financial plan<br />
targets <strong>and</strong> writing the monthly report.<br />
18.4.1 Automation in the brewhouse<br />
In this increasingly common situation automation of the sequence of brewhouse<br />
operations is indispensable. This automation is no substitute for the brewhouse operator<br />
knowing what is happening in the vessels <strong>and</strong> his training in this respect is crucial. Very<br />
expensive mistakes can ensue if automation is slavishly followed when things are going<br />
wrong. The big change in automation in the last 20 years has been the gradual replacement<br />
of hard-wired relay logic systems with the silicon chip based programmable logic<br />
controllers. This has been coupled with the very rapid development of personal computers<br />
(PCs), which has provided user-friendly operator interfaces <strong>and</strong> the ability to store huge<br />
amounts of data. Concurrent with all this have been big reductions in the capital cost of<br />
equipment, as the `Microsoft' operating systems have come to dominate the world. The<br />
developments in PCs have facilitated <strong>and</strong> made cheaper industrial systems' development<br />
in what has truly been a buyer's market. A major problem is that systems have advanced so<br />
fast that obsolescence has become a factor that the brewer cannot ignore. The language of<br />
the control systems expert has also become difficult <strong>and</strong> obscure for the brewing<br />
technologist <strong>and</strong> this has sometimes led to increased expense when supplier <strong>and</strong> customer<br />
have not understood one another. The old adage, `the customer is always right' still holds<br />
but it is imperative in this area that the customer sets out very precisely what is required of<br />
the system he is buying. There is no substitute for the `Design Brief' <strong>and</strong> the `User<br />
Requirement Specification', which must form the basis of the Contract. The brewer (the<br />
customer) should remain the most important person in the deal.<br />
Systems are usually designed bespoke <strong>and</strong> are not universal. However it is possible to<br />
make some comment on the most successful systems. The way the system is put together<br />
is called architecture, which describes the links through from plant level to a management<br />
information system (MIS). Several important steps can be recognized.<br />
Sensors<br />
The success of the system, of course, starts with measurements. No amount of clever<br />
software will make up for an inaccurate or imprecise measurement. The quality of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
initial measuring is paramount. Fortunately, in the brewhouse we usually want to measure<br />
temperature, pH value <strong>and</strong> pressure <strong>and</strong> equipment for these measurements is cheap <strong>and</strong><br />
reliable.<br />
Programmable logic controllers (PLCs)<br />
The control at plant level is effected through PLCs. The PLC is a computer that is<br />
programmed to h<strong>and</strong>le simple input <strong>and</strong> output instructions, e.g., `open', `close'. Inputs<br />
are the signals from sensors <strong>and</strong> outputs are the instructions to the machines; pumps,<br />
motors, valves, etc. Normally a brewery company will specify one type of PLC from one<br />
manufacturer <strong>and</strong> this is highly desirable for consistency over the whole brewery site <strong>and</strong><br />
for future changes to the system.<br />
Supervisory control <strong>and</strong> data acquisition systems (SCADA)<br />
PLCs link machines together <strong>and</strong> send out simple instructions to make changes in the<br />
process according to how they have been programmed. The operator requires information<br />
on what is happening <strong>and</strong> needs to store information for future use. This is provided<br />
through a SCADA system that can communicate with PLCs by an industrial Ethernet.<br />
The SCADA system can now have a very user-friendly interface through a PC (which can<br />
use familiar Microsoft Windows NT software) in the brewhouse. The SCADA system can<br />
also be linked to high-level MIS that can further be linked nationally or even<br />
internationally providing Head Office with direct access to an individual brewery.<br />
It is now possible to use this system architecture to hold the recipes of the brewing<br />
process <strong>and</strong> to schedule the brewhouse operations (Cooper et al., 2002). A series of<br />
reports can be obtained that could detail actual achieved mash temperatures, transfer<br />
times <strong>and</strong> run-off profiles, etc. Studying these reports allows predictions to be made <strong>and</strong><br />
correlations to be sought which can lead to future brewhouse improvements.<br />
All these developments make the operation of the brewhouse easier to control in these<br />
days of very low manpower levels. It must again be stressed how important the training<br />
of the operator is so that timely intervention may be made when indicated by the control<br />
system. The development of the qualification the `General Certificate in <strong>Brewing</strong> <strong>and</strong><br />
Packaging' (formerly the Foundation Certificate) by the Institute <strong>and</strong> Guild of <strong>Brewing</strong> is<br />
a particularly important step in this respect (Brookes, 1999).<br />
18.4.2 Scheduling of brewhouse operations<br />
The brewhouse is the most important part of the brewery in relation to its overall<br />
capacity. The number of brews that can be carried out in one day <strong>and</strong> the length of the<br />
brewing week are the relevant factors. Different businesses will have very different<br />
requirements. The driving force for the micro- or pub brewery will be to have the beer<br />
presented to the customer in the pub in the freshest possible state. There will be<br />
seasonality here <strong>and</strong> the micro-brewer will need to be able to increase production in the<br />
summer to meet dem<strong>and</strong>. However it is unlikely that there will be more than two or three<br />
brews performed in one day <strong>and</strong> the brewing week will probably be five days but those<br />
days may include the weekend so as to ensure maximum exposure to the public if<br />
brewing vessels are on view. This will give two days for thorough cleaning which is<br />
essential to maintaining product quality.<br />
The situation is very different in the large brewery producing national or international<br />
br<strong>and</strong>s. It is now frequently the case that beer will not be delivered direct to the customer<br />
from the brewery but will go to a regional distribution centre where stock will be held<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(Chapter 22). The brewery's staff often feels remote from the customer in this situation.<br />
The brewery's success will be measured against its ability to produce beer to aplan <strong>and</strong><br />
this plan will be derived from historical sales data <strong>and</strong> marketing forecasts. The plan will<br />
often be given to the brewery from acentral planning department who will monitor<br />
progress. The role of the brewery is clear. It must deliver the plan.<br />
The core of the plan then will be the number of brews that can be consistently<br />
delivered from the brewhouse. Working arrangements in the brewhouse will probably be<br />
based on some form of 168-hour cover where teams will cover the whole week <strong>and</strong> take<br />
rest days on a rotating basis. The weekends are not significant since any overtime<br />
payments are likely to be built in to an overall salary for the job. This makes scheduling<br />
easier. However, it is still very important to build in to the programme sufficient time for<br />
cleaning,maintenance <strong>and</strong>calibrationofplant<strong>and</strong>itisunlikelythattheplan willdem<strong>and</strong><br />
more than six-day brewing. There will also be time needed for maintenance <strong>and</strong> six-day<br />
brewing cannot be sustained indefinitely. Trade is usually such that there is considerable<br />
seasonality in dem<strong>and</strong> with peak brewing being required in the summer <strong>and</strong> at national<br />
holiday periods or periods of Company sponsorship of major events such as football<br />
championships. These external factors become the driving force to the plan <strong>and</strong> influence<br />
the calculation of brewhouse capacity. The maximum capacity of the brewhouse will be<br />
the number of brews in aday length of brewing week number of weeks of brewing<br />
in ayear. To be prudent in calculating seasonality we might express the foregoing as 10<br />
6 46 ˆ2,760 brews per year. The planning department would build this information<br />
into its model of operations <strong>and</strong> hence the production <strong>and</strong> distribution plan would be<br />
derived.<br />
The actual volume of beer produced depends, of course, on the strength of wort<br />
delivered from the brewhouse <strong>and</strong> the amount of post-fermentation dilution carried out.<br />
This relates to the types of beer being brewed <strong>and</strong> the nature of the grists. Scheduling of<br />
brewhouse operations will usually include scheduling the intake of raw materials <strong>and</strong> this<br />
will also frequently be arranged centrally in large brewing groups. To service alarge<br />
brewery as illustrated above, considerable quantities of raw materials will require to be<br />
delivered each week:<br />
Let us assume that each brew is 1,000 hl ( 600 imp. brl) at agravity of 1048ë<br />
(48ëSacch, 12ëP). Let us also assume that the grist is 90% white malt <strong>and</strong> 10% sugar<br />
(Section 18.5). Each brew will require approximately 15 tonnes of malt <strong>and</strong> 1.5 tonnes of<br />
sugar solution. For asix-day week at ten brews per day this amounts to 900 tonnes of<br />
malt <strong>and</strong> 90 tonnes of sugar per week therefore we require around 36 loads of malt <strong>and</strong> 4<br />
loads of sugar solution per week.<br />
This represents aconsiderable schedulingoperation<strong>and</strong> requiresclose liaisonbetween<br />
the maltster, the sugar supplier <strong>and</strong> the brewer. To be prudent the brewer will probably<br />
want to hold at least one week's stock of raw materials on site <strong>and</strong> so will require at least<br />
ten 100-tonne malt bins <strong>and</strong> three to four 20-tonne sugar tanks. Of course there may also<br />
be arequirement for delivery <strong>and</strong> storage of smaller quantities of special malts <strong>and</strong> mash<br />
tun adjuncts (see Chapter 5).<br />
In recent years this whole subject has received much attention from planners <strong>and</strong><br />
software experts <strong>and</strong> computer systems with the general title of materials requirements<br />
planning (MRP) <strong>and</strong> the further development of manufacturing reserve planning (MRPII)<br />
are now commonplace. The objective is to pick up the principles of `just in time' (JIT)<br />
production to minimize the stock holding of raw materials, those in-process <strong>and</strong> finished<br />
beer in the brewery. This in turn reduces working capital. Software systems in use include<br />
PRMS, PRISM <strong>and</strong> SAP. These systems can be extremely useful particularly in providing<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
information to suppliers of raw materials. But they are no substitute for physical stock<br />
checks in the brewery to ensure that the materials required <strong>and</strong> apparently in stock are<br />
actually present!<br />
18.5 Economic aspects of brewhouses<br />
Fixed costs in breweries relate to manpower, utilities, repairs <strong>and</strong> depreciation <strong>and</strong> other<br />
costs such as rates <strong>and</strong> other local taxes. Variable costs relate to the efficiency in the use<br />
ofmaterials <strong>and</strong> losses of product incurred during processing. Of prime importance tothe<br />
efficiency of the brewhouse then, <strong>and</strong> hence to brewery variable cost, is the quantity of<br />
soluble material extracted from the grist. The more efficient this process the greater will<br />
be the contribution to the fixed costs of running the brewery, where contribution is<br />
defined as the income less the variable costs (sometimes called gross profit).<br />
The strength of wort is, of course, judged by the quantity of material in solution, the<br />
extract. This is usually measured as the specific gravity relative to water at aspecific<br />
temperature (see also Chapter 4), or by relating to the concentration (%w/w) of asucrose<br />
solution having the same specific gravity using the scales of Balling or Plato. A pale malt<br />
ground <strong>and</strong> extracted in cold water will yield 16±22% of its dry matter as soluble extract<br />
(the cold water extract, Chapter 4). The same malt ground <strong>and</strong> extracted at 65 ëC (149 ëF)<br />
will yield 75±83% of its dry matter as extract (the hot water extract). The extra substances<br />
yielded by hot water mashing derive from enzymic attack on initially insoluble materials,<br />
mainly starch. This forms the nub of potential economic improvement in the brewhouse.<br />
If increases in soluble extract can be consistently achieved then there will be reductions<br />
in the raw material bill, lower variable cost <strong>and</strong> brewhouse operations will be judged a<br />
financial success. Consider the preparation of 1,000 hl of wort at a gravity of 1,048 ë<br />
(48 ëSacch, 12 ëP), from a grist of 90% white malt <strong>and</strong> 10% sugar:<br />
Let the extract of the white malt be 300 l ë/kg (as is). Let the extract of the sugar<br />
solution be 320 l ë/kg (as is). Let the extract efficiency of the brewhouse be 97% (i.e. in<br />
normal circumstances, 97% of the extract as indicated by laboratory analysis would be<br />
expected to be extracted into brewery wort). The amount <strong>and</strong> strength of wort to be<br />
collected is:<br />
1000 48 ˆ 48; 000 hl …4;800;000 l †<br />
the quantity of materials required will be:<br />
white malt …4,800,000 0:9† 0:97 300 ˆ 14,845 kg<br />
sugar …4,800,000 0:1† 320 ˆ 1,500 kg<br />
The sugar is not affected by the 97% extract recovery because it is added directly to<br />
the wort boiler <strong>and</strong> is assumed to be utilized at 100%. Clearly, if the extract efficiency<br />
can be consistently raised to 98% then 1% less malt will be required, i.e., 148 kg of malt<br />
less. Over the course of 2,000 brews in a year this would amount to around 300 tonnes of<br />
malt, which could equate to variable cost savings of around US $96,000 per year. This is<br />
a considerable sum <strong>and</strong> illustrates the great importance of striving to achieve consistent<br />
improvement in brewhouse performance by achieving operating conditions to optimize<br />
enzyme activity.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
18.6 Summary<br />
Brewhouses are the heart of every brewery whether large multi-national or niche br<strong>and</strong><br />
pub brewery. It is the extraction of the grist that is crucial to economic success <strong>and</strong> to<br />
yeast performance <strong>and</strong> beer quality. Brewhouses can be of different types <strong>and</strong> large<br />
brewhouses are capable of being operated with high efficiency <strong>and</strong> consistency whilst<br />
small brewhouses are capable of being controlled to yield worts for beers with unique <strong>and</strong><br />
stronger flavours. Further improvements are likely to be restricted by the amount of<br />
money now being invested in basic research. In the long run this will be a short-sighted<br />
policy. The long-term winners will be those organizations that invest in studying how<br />
enzymes cope best with the heterogeneous conditions of the brewery mash.<br />
18.7 References<br />
ANDERSON, R. G. (1993) Ferment, 6, 191.<br />
BAETSLE, G. (1983) Cerevisia, 7, 11.<br />
BRIGGS, D. E., HOUGH, J. S., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1981) Malting <strong>and</strong> <strong>Brewing</strong> Science Volume 1,<br />
2nd edn, Chapman <strong>and</strong> Hall, London <strong>and</strong> New York.<br />
BROOKES, P. A. (1999) Proc. 7th. Conv. Inst. Brew. (Africa Section), Nairobi, xiii.<br />
COOPER, T. J., BARNES, Z. C. <strong>and</strong> MCFARLANE, I. K. (2002) Proc. 27th Conv. Inst. Brew. (Asia Pacific<br />
Section), Adelaide, 1.<br />
DE CLERK, J. <strong>and</strong> DE CLERK, E. (1965) Tech. Quart. MBAA., 2, 183.<br />
JEFFERY, E. J. (1956) <strong>Brewing</strong>, Theory <strong>and</strong> Practice, Nicholas Kaye, London, 18.<br />
LEWIS, M. J. <strong>and</strong> LEWIS, D. J. (1996) Proc. 6th Int. Brew. Tech. Conf., Harrogate, 227.<br />
MALONE, R. (2001) Brewers' Guard., 130 (9), 24.<br />
MOLL, M. <strong>and</strong> MIDOUX, N. (1985) Tech. Quart. MBAA., 22, 67.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
19<br />
Chemical <strong>and</strong> physical properties of beer<br />
19.1 Chemical composition of beer<br />
Beer, the final product of the brewing process, is designed to be drunk. It is acomplex<br />
mixture; well over 450 constituents have been characterized, <strong>and</strong>, in addition, it contains<br />
macromolecules such as proteins, nucleic acids, polysaccharides <strong>and</strong> lipids. Together all<br />
theseconstituentsproducethecharacterofbeer.Inthischapterwereviewthecompounds<br />
that have been found in beer but delay discussion of the influence they have on beer<br />
flavour until the next chapter. However, to prevent duplication, data on the taste<br />
thresholds of the compounds found are included in the Tables given in this Chapter. A<br />
useful introduction is given in the monograph Beer: Quality, Safety <strong>and</strong> Nutritional<br />
Aspects (Baxter <strong>and</strong> Hughes, 2001). Jurado has reviewed beer styles <strong>and</strong> provided<br />
analyses for Pilsener (2002a), Munich Helles (2002c), brown ales (2001a), strong ales<br />
(2002e), pale ales (incl. IPA) (2002e), red beers (2002b), wheat beers (2001b) <strong>and</strong><br />
seasonal <strong>and</strong> special beers (2002d). The ranges of analytical values found are collected in<br />
Table 19.1. A. Piendl (Weihenstephan) has published analyses of beers from all over the<br />
world. In most cases the units used by the authors quoted have been retained. It is often<br />
not known whether parts per million (ppm) is mg/l or mg/kg. Parts per billion (ppb) is<br />
parts per 10 9 ( g/l or g/kg).<br />
Beer constituents can be divided into volatile <strong>and</strong> non-volatile components. The<br />
volatile components have greater vapour pressure <strong>and</strong> are responsible for the bouquet or<br />
aroma of beer. They are concentrated in the headspace above the liquid in a closed<br />
container <strong>and</strong> will pass into the distillate if the beverage is distilled. The complex mixture<br />
of volatile components either in the headspace or in a solvent extract of the beer can be<br />
resolved by gas-liquid chromatography, using either packed or capillary columns, <strong>and</strong> the<br />
components identified by mass spectrometry (GC-MS). The non-volatile constituents<br />
include inorganic salts, sugars, amino acids, nucleotides, polyphenols <strong>and</strong> hop resins<br />
together with macromolecules such as polysaccharides, proteins <strong>and</strong> nucleic acids. Such<br />
compounds are usually resolved by high precision liquid chromatography (HPLC).<br />
Before the advances in chromatography available today, brewers developed numerous<br />
methods of analysis to control their process <strong>and</strong> the quality of their products <strong>and</strong> many of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.1 Beer analyses<br />
100% Malt Munich-style Brown Strong IPA Pale ales Red beers Wheat beers Seasonals<br />
Pilsener Helles ales ales <strong>and</strong> specials<br />
Ref. Jurado Jurado Jurado Jurado Jurado Jurado Jurado Jurado Jurado<br />
(2002a) (2002c) (2001a) (2002e) (2002e) (2002e) (2002b) (2001b) (2002d)<br />
No. of samples 18 33 22 8 17 58 23 27 34<br />
Specific 1.00682± 1.00258± 1.00599± 1.00676± 1.00739± 1.00471± 1.00627± 1.00449± 1.00342±<br />
gravity 1.01411 1.10166 1.01991 1.02715 1.01646 1.01680 1.01986 1.02385 1.03471<br />
(1.009740) (1.01485) (1.0132) (1.01236) (1.01326)<br />
Apparent 1.75±3.60 0.66±3.31 1.54±5.06 1.73±6.86 1.89±4.19 1.21±4.83 1.61±5.04 1.15±6.04 1.57±8.71<br />
extract (2.48) (2.04) (3.36) (3.15) (2.67)<br />
Alcohol 3.05±4.66 3.26±4.81 2.11±5.02 4.02±8.00 2.25±7.26 2.93±5.89 3.17±6.28 3.03±6.51 1.46±5.63<br />
(% w/w) (4.30) (4.11) (3.97) (3.88) (4.04)<br />
Real extract 3.52±5.17 2.64±4.89 3.58±6.88 4.65±10.24 3.31±6.92 2.80±6.83 3.31±6.77 2.91±7.46 2.75±9.37<br />
(% w/w) (4.30) (3.92) (5.20) (4.93) (4.53)<br />
Original 10.87±13.06 9.88±13.50 7.77±15.5 12.43±24.50 7.78±20.35 8.99±17.50 9.81±18.42 10.01±19.32 8.0±17.0<br />
gravity ( o P) (11.97) (11.89) (12.855) (12.42) (12.34)<br />
Real degree 56.0±70.2 59.2±76.4 52.1±71.2 56.8±74.8 58.2±70.8 53.4±76.6 52.7±71.0 45.6±72.9 24.3±75.9<br />
of fermentation (65.4) (68.4) (61.194) (62.0) (65.0)<br />
Calories 143.6±172.0 63.5±178.6 102.2±209.8 165.0±346.6 124.2±278.5 125.7±238.5 127.6±250.8 130.2±264.8 106.0±230.5<br />
(158.0) (156.2) (171.9) (165.7) 163.7)<br />
pH 3.74±4.63 3.99±4.77 3.95±4.56 3.81±4.83 3.87±4.74 3.78±4.64 3.86-4.99 3.66±4.82 3.14±4.60<br />
(4.33) (4.37) (4.33) (4.21) (4.46)<br />
Colour 2.9±8.8 3.1±15.3 18.4±58.9 14.9±47.4 6.7±26.0 4.1±52.5 9.4±54.4 3.8±49.0 4.4±68.7<br />
(Lovibond) (6.0) (4.2) (32.8) (23.9) (11.2)<br />
Bitterness 3.1±51.2 13.7±30.5 13.5±44.7 15.4±67.3 24.3±78.6 6.5±67.2 12.4±35.5 8.2±33.9 7.5±42.2<br />
(IBU) (31.0) (18.4) (26.6) (21) 17.2)<br />
Vicinal diketones (ppm) 0.01±0.04 0.01±0.09 0.01±0.25 0.03±0.29 0.01±0.20 0.01±0.22 0.01±0.16 0.01±0.38 0.01±0.22<br />
(0.0) (0.04) (0.04) (0.04) (0.05)<br />
Sodium (Na, ppm) 3±45 11±50 21±106 24±322 14±113 13±123 17±100 14±86 13±150<br />
(34) (26) (39) (43) (43)<br />
Dimethyl sulphide (ppb) 3±73 ± ± ± ± ± ± ± ±<br />
(50±55)<br />
SO2 (ppm) ± 0±3.2 ± ± ± ± ± ± ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.2 Analyses of British beers (Anon., 1960, 1967)<br />
Quality OG Alcohol Unfermented Isohumulones<br />
(% v/v) matter (%) (mg/l)<br />
Draught bitter 1030.9±1045.3 3.0±4.6 27±45 20±40<br />
Draught mild 1030.7±1036.5 2.5±3.6 29±48 14±37<br />
Light ale (bottle or can) 1030.6±1038.9 2.9±4.0 30±40 16±38<br />
Best pale ale 1040.3±1050.3 4.3±6.6 21±43 19±55<br />
Brown ale 1030.2±1040.6 2.5±3.6 43±55 16±28<br />
Stout ± Guinness 1040.0±1046.1 4.4±5.1 30 55±62<br />
± Mackeson 1044.3±1047.6 3.7±3.8 49 27±31<br />
Strong ales 1065.9±1077.7 6.1±8.4 32±44 25±43<br />
Lagers 1029.7±1036.3 3.3±3.6 35±39 20±32<br />
these are retained in Analytica-EBC <strong>and</strong> by the Institute of <strong>Brewing</strong> <strong>and</strong> the American<br />
Society of <strong>Brewing</strong> Chemists (Section 1.15.1, p. 9). Some representative analyses are<br />
given in Tables 19.1 <strong>and</strong> 19.2.<br />
Another way to classify the organic constituents of beer is with reference to the<br />
heteroatoms present. Beers contain only trace amounts of hydrocarbons, the majority of<br />
the constituents contain carbon, hydrogen <strong>and</strong> oxygen. Small amounts of nitrogencontaining<br />
constituents are present of which the proteins are important for the physical<br />
properties of beer. Only low levels of sulphur-containing compounds are found but<br />
volatile sulphur compounds have low thresholds <strong>and</strong> so small amounts can influence the<br />
flavour of beer. Indeed, abnormally high levels of sulphur compounds can be responsible<br />
for off-flavours. Volatile sulphur compounds can be examined by gas chromatography<br />
usingeitheraflame photometricdetectororaSievers'chemiluminescentdetector bothof<br />
which are specific for sulphur-containing compounds. In general different beers contain<br />
different proportions of the same constituents. Only when novel raw materials are used<br />
will novel constituents be found in the beer as, for example, Belgian fruit beers such as<br />
Framboise. However, accidental or deliberate contamination of wort or beer with foreign<br />
micro-organisms may well produce new metabolites <strong>and</strong> flavours.<br />
19.1.1 Inorganic constituents<br />
The most abundant constituent of beer is water, the medium in which, in bright beer, all<br />
theotherconstituentsaredissolved.<strong>Brewing</strong>liquornormallycontainsonlytraceamounts<br />
of organic matter <strong>and</strong> the desirable cations <strong>and</strong> anions required in the liquor are reviewed<br />
inChapter3.Duringthebrewingprocessothersaltswillbeextractedfrommalt<strong>and</strong>hops,<br />
some ions may be precipitated on the break <strong>and</strong> others may be absorbed by the yeast so<br />
that the inorganic salts present in beer are very different from those in the liquor used.<br />
Data for the inorganic constituents of beer are collected in Table 19.3.<br />
The major ions are the cations potassium, sodium, calcium <strong>and</strong> magnesium <strong>and</strong> the<br />
anions chloride, sulphate, nitrate <strong>and</strong> phosphate. There are agreed international methods<br />
for measuring sodium, potassium, calcium <strong>and</strong> magnesium in beer by atomic absorption<br />
spectroscopy. In addition the ASBC gives methods for determining calcium <strong>and</strong><br />
magnesiumbytitrationwithEDTA.Internationalmethodsarealsoavailabletodetermine<br />
the anions in beer, chloride, sulphate, nitrate <strong>and</strong> phosphate by ion chromatography.<br />
There is also an international method for chloride by conductometry, <strong>and</strong> Analytica-EBC<br />
gives agravimetric method for sulphate <strong>and</strong> an enzymatic method for nitrate. Excess<br />
nitrate in beer is undesirable as potentially nitrates can be reduced to nitrites, which with<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.3 Inorganic constituents of beer<br />
Constituent Source Concentration Reference<br />
range mg/l (ppm)*<br />
Aluminium German 0.1±1.24 Postel et al. (1983)<br />
Arsenic Lagers 0.02 Binns et al. (1978)<br />
Spanish 3.1±8.2 Cervera et al. (1989)<br />
Others 1.8±11.2 Cervera et al. (1989)<br />
Cadmium German 0.0002±0.020 Postel et al. (1983)<br />
Spanish 0.031±0.397 YbaÂnÄez et al. (1989)<br />
Others 0.095±0.677 YbaÂnÄez et al. (1989)<br />
Calcium British 40±140 Paul <strong>and</strong> Southgate (1978)<br />
German 3.8±102 (32.7) Postel et al. (1974)<br />
Lagers 10±135 (36) Binns et al. (1978)<br />
Chromium European (0.0072) Robberecht et al. (1984)<br />
Spanish 0.004±0.022 Farre et al. (1987)<br />
Cobalt Spanish 0.00005±0.00079 YbaÂnÄez et al. (1989)<br />
Others 0.00005±0.00038 YbaÂnÄez et al. (1989)<br />
Copper British 0.3±0.8 Paul <strong>and</strong> Southgate (1978)<br />
German 0.04±0.80 (0.19) Postel et al. (1972b)<br />
0.02±1.55 Postel et al. (1983)<br />
Spanish 0.0064±0.0603 (0.029) YbaÂnÄez et al. (1989)<br />
Others 0.010±0.040 (0.027) YbaÂnÄez et al. (1989)<br />
Lagers 0.01±0.41 (0.11) Binns et al. (1978)<br />
Iron British 0.1±0.5 Paul <strong>and</strong> Southgate (1978)<br />
German 0.02±0.84 (0.02) Postel et al. (1972a)<br />
0.04±1.55 Postel et al. (1983)<br />
Lagers 0.04±0.44 (0.12) Binns et al. (1978)<br />
Wheat beer (0.63) Postel et al. (1972a)<br />
Lead German 0.003±0.024 Postel et al. (1983)<br />
Spanish 0.0014±0.0056 (0.0028) YbaÂnÄez et al. (1989)<br />
Others 0.0008±0.0025 (0.0017) YbaÂnÄez et al. (1989)<br />
Lagers 0.06 Binns et al. (1978)<br />
Magnesium British 60±200 Paul <strong>and</strong> Southgate (1978)<br />
German 75±250 (114) Postel et al. (1974)<br />
Lagers 34±162 (82) Binns et al. (1978)<br />
Manganese German 0.04±0.51 (0.20) Postel et al. (1973)<br />
Mercury 0±0.0008 Donhauser et al. (1987)<br />
Nickel 0±0.26 Brenner et al. (1965)<br />
Phosphorous British 90±400 Paul <strong>and</strong> Southgate (1978)<br />
Australian 96±304 (196) Bottomley <strong>and</strong> Lincoln (1958)<br />
Potassium British 330±1100 Paul <strong>and</strong> Southgate (1978)<br />
German 396±562 (476) Kieninger (1978)<br />
Mexican 220±358 Canales et al. (1970)<br />
Lagers 253±680 (362) Binns et al. (1978)<br />
Selenium 0±0.0072 (0.0012) Donhauser et al. (1987)<br />
Silicon 10.2±22.4 Anderson et al. (1995)<br />
Sodium British 40±230 Paul <strong>and</strong> Southgate (1978)<br />
German 9±120 (35) Kieninger (1978)<br />
Lagers 15±170 (58) Binns et al. (1978)<br />
Tin German 0.010±0.020 Postel et al. (1983)<br />
Zinc German 0.01±1.48 (0.10) Postel et al. (1975)<br />
Lagers 0.01±0.46 Binns et al. (1978)<br />
German 0.01±0.26 Postel et al. (1983)<br />
ANIONS<br />
Chloride British 150±984 Paul <strong>and</strong> Southgate (1978)<br />
German 143±365 (210) Kieninger (1978)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.3 Continued<br />
Constituent Source Concentration Reference<br />
range mg/l (ppm)*<br />
Fluoride British 0.08±0.71 Warnakulasuriya et al. (2002)<br />
German 0.08±0.64 (0.15) Postel et al. (1976)<br />
Nitrate German 1.4±101.3 (34.0) Postel (1976)<br />
German 13±43 Gmelch et al. (1989)<br />
Phosphate British 260±400 Paul <strong>and</strong> Southgate (1978)<br />
German 624±995 (860) Kieninger (1978)<br />
Sulphate British 150±400 Paul <strong>and</strong> Southgate (1978)<br />
German 107±398 (182) Kieninger (1978)<br />
African lager 125±260 Shah (1975)<br />
* Average values in parentheses.<br />
amines can form carcinogenic N-nitrosamines (see later). The EEC limit for nitrates in<br />
drinking water is 25 mg/litre.<br />
Trace amounts of many metals are essential for yeast growth whereas larger amounts<br />
may be toxic <strong>and</strong> may be limited by law. When specific limits for beer are not specified<br />
the limits for potable water are usually applied. In Britain, the levels of arsenic <strong>and</strong> lead<br />
are limited to 0.2 mg/kg (ppm) <strong>and</strong> the Institute of <strong>Brewing</strong> describes methods for<br />
ensuring these limits are met. International methods for determining iron, copper <strong>and</strong> zinc<br />
by atomic absorption spectroscopy are published as well as spectrophotometric methods<br />
for iron <strong>and</strong> copper. The UK Food St<strong>and</strong>ards Committee recommends a limit for copper<br />
in beer of 7.0 mg/kg. Nickel can be estimated in beer spectrophotometrically by<br />
measuring the colour formed with dimethylglyoxime (Analytica-EBC). In the 1960s<br />
cobalt was added to beer to improve foam stability <strong>and</strong> prevent gushing. However, this<br />
was found to cause acute heart disease in heavy drinkers (in excess of 20 pints beer/day)<br />
so the <strong>practice</strong> was discontinued (Long, 1999).<br />
Carbon dioxide is a natural product of fermentation <strong>and</strong> beers contain 3.5±4.5 g/l.<br />
Supersaturated beers <strong>and</strong> naturally conditioned beers may contain as much as 6 g/l but gas<br />
will be evolved as soon as the pressure is released. The sensory threshold of carbon<br />
dioxide is about 1 g/l so the amount present will influence the flavour of beer. Analytica-<br />
EBC describes a titrimetric <strong>and</strong> an instrumental method for measuring the amount of CO2<br />
present. The Institute of <strong>Brewing</strong> describes a manometric method <strong>and</strong> the ASBC gives<br />
two methods, one for beer in tanks <strong>and</strong> the other for beer in bottles or cans.<br />
19.1.2 Alcohol <strong>and</strong> original extract<br />
In the European Economic Community (EEC) the strength of beer <strong>and</strong> other alcoholic<br />
drinks, is expressed as alcohol by volume (ABV), that is the ratio of the volume of the<br />
ethyl alcohol contained in the liquor to the volume of the liquor including the ethyl<br />
alcohol (expressed as a percentage to one decimal place). Further the EEC directed in<br />
1987 that the ABV `alc x.y% vol' shall appear on the label. In many countries excise duty<br />
is levied on the basis of the ABV. In Britain duty is payable on beers with more than<br />
1.2% ABV. Before 1993 duty in Britain was calculated from the original gravity of the<br />
wort fermented. Beers can contain between 0.05% ABV, in an alcohol-free beer, up to<br />
about 12.5%. In Germany bock beers must contain more than 6.0% ABV <strong>and</strong> doublebock<br />
beers more than 7.5% ABV. According to Glaser (2002), the strongest beer is<br />
Utopias MM II with 24% ABV. Analytica-EBC, the Institute of <strong>Brewing</strong> <strong>and</strong> the ASBC<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
provide several methods for measuring alcohol in beers but national Excise authorities<br />
may specify their own modifications.<br />
Thedistillationmethodisusuallyregardedasthereferencemethodtobeusedinanycases<br />
ofdispute.Hereanaccuratelymeasuredquantityoffilteredbeer(say100.0mlinavolumetric<br />
flask at 20ëC or 100.00g) is washed into asuitable distillation flask (400±500ml capacity)<br />
<strong>and</strong> distilled, taking care not to char the residue in the distillation flask, until c. 85ml is<br />
collected (inthesame volumetricflask). Thisisthenmadeupto100.00ml(or100.00g)<strong>and</strong><br />
thespecificgravityinairat20ëC/20ëCismeasuredinasuitablepycnometer(Reischauer)or<br />
specificgravitybottletofiveplacesofdecimals.APaardensitometerorasimilarinstrument,<br />
whichmakethismeasurementelectronicallyisnowcommonlyused(seeIoB,ASBC,Mundy,<br />
1996). Several tables exist to convert the specific gravity of the distillate to give the alcohol<br />
content.TheEECrecommendthatthetablesoftheOrganisationInternationaledeMetrologie<br />
Legale (OIML) should be used but they refer to specific gravity in vacuo. The EBC has<br />
therefore provided an alcohol table based on the specific gravity measured in air <strong>and</strong> give<br />
polynomials for alcohol <strong>and</strong> extract (Rosendahl <strong>and</strong> Schmidt, 1987):<br />
A% m=m ˆ517:4…1 SGA†‡5084…1 SGA† 2 ‡33503…1 SGA† 3<br />
where Ais %alcohol by weight <strong>and</strong> SGA is the specific gravity of the distillate in air at<br />
20ëC/20ëC.<br />
To convert A% m/m to A% v/v (ABV):<br />
Alcohol %v=v ˆ<br />
A%m=m SGA<br />
0:791<br />
where 0.791 is the specific gravity of ethanol at 20ëC/20ëC.<br />
The American Society of <strong>Brewing</strong> Chemists <strong>and</strong> the British Excise Authorities<br />
provide their own tables (for part of the latter see Table 19.4).<br />
Alcohol may also be determined by catalytic combustion using a Servochem<br />
Automatic Beer Analyser (SCABA). The injected beer is divided into two streams, one<br />
enters a Paar U-tube densitometer, the other passes down a column as a falling film where<br />
the alcohol is removed as a vapour with a countercurrent of air <strong>and</strong> passed over an<br />
alcohol sensor. After calibration with known st<strong>and</strong>ards, the onboard computer will<br />
display the % alcohol either m/m or v/v when the results are found to agree with the<br />
distillation method (Freeston <strong>and</strong> Baker, 1993). Schropp et al. (2002) evaluated an NIR<br />
procedure for measuring the alcohol content as in the Alcolyzer instrument.<br />
The amount of alcohol present can also be determined from the refractive index of the<br />
media. This is much quicker than the distillation method but requires st<strong>and</strong>ardizing to<br />
determine the constants A, B, <strong>and</strong> C in the regression equation:<br />
Alcohol % …v=v† ˆ A…SR WR† B…PG† C<br />
where SR is the refractive index of the beer, WR is the refractive index of water, <strong>and</strong> PG<br />
is the present gravity of the beer {1000 (SG 1.00000)}. The instrument is thermostatted<br />
<strong>and</strong> measurements are usually made at 20 ëC. Ethanol can also be determined by gas<br />
chromatography using a flame ionization detector <strong>and</strong> direct injection on to a suitable<br />
column (Poropak Q, 15% Carbowax 20 M, SGE BP20 or CP Wax 52 CB), after a known<br />
addition of n-butanol as an internal st<strong>and</strong>ard. The amount of alcohol is calculated by<br />
comparison of the peak areas (see also Clarkson et al., 1995).<br />
For alcohol-free or low-alcohol beers (< 0.008%) an enzymatic method is given based<br />
on the Boehringer test kit. The alcohol is oxidized first to ethanal <strong>and</strong> then to ethanoic<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.4 Alcohol table giving the percentage of ethanol (v/v) at 20 ëC from the density of solution in air (kg/cu m) at 20 ëC. (Crown Copyright)<br />
0 1 2 3 4 5 6 7 8 9<br />
788.16 100.0 ± ± ± ± ± ± ± ± ±<br />
780 ± ± ± ± ± ± ± ± ± 99.84<br />
790 99.64 99.44 99.24 99.04 98.83 98.63 98.40 98.18 97.96 97.74<br />
800 97.51 97.28 97.05 96.81 96.57 96.33 96.09 95.84 95.59 95.33<br />
810 95.08 94.82 94.56 94.30 94.03 93.76 93.49 93.22 92.94 92.63<br />
820 92.38 92.09 91.80 91.52 91.22 90.93 90.63 90.33 90.03 89.73<br />
830 89.42 89.12 88.81 88.49 88.18 87.86 87.55 87.23 86.90 86.58<br />
840 86.25 85.53 85.60 85.26 84.93 84.59 84.24 83.92 83.58 83.23<br />
850 82.89 82.54 82.20 81.85 81.50 81.14 80.79 80.43 80.08 79.72<br />
860 79.35 78.99 78.63 78.26 77.89 77.53 77.16 76.78 76.40 76.03<br />
870 75.66 75.28 74.90 74.51 74.13 73.75 73.36 72.97 72.58 72.19<br />
880 71.79 71.40 71.00 70.60 70.20 69.80 69.39 68.99 68.58 68.17<br />
890 67.76 67.34 66.93 66.51 66.09 65.67 65.25 64.83 64.40 63.97<br />
900 63.54 63.11 62.68 62.24 61.80 61.36 60.92 60.47 60.03 59.58<br />
910 59.13 58.67 58.22 57.76 57.30 56.84 56.37 55.90 55.43 54.96<br />
920 54.48 54.00 53.52 53.03 52.54 52.05 51.55 51.05 50.55 50.04<br />
930 49.53 49.01 48.49 47.97 47.44 46.91 46.36 45.82 45.26 44.71<br />
940 44.15 43.58 43.00 42.41 41.82 41.23 40.61 39.99 39.37 38.73<br />
950 38.08 37.42 36.75 36.07 35.37 34.67 33.95 33.21 32.47 31.69<br />
960 30.92 30.12 29.32 28.48 27.55 26.80 25.92 25.04 24.14 23.24<br />
970 22.32 21.40 20.47 19.54 18.60 17.68 16.74 15.84 14.93 14.03<br />
980 13.14 12.27 11.40 10.56 9.72 8.90 8.09 7.30 6.52 5.76<br />
990 5.01 4.27 3.54 2.83 2.13 1.44 0.77 0.10 ± ±<br />
997.15 0.00<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
acid with nicotinamide adenine dinucleotide (NAD ‡ ) <strong>and</strong> the reduction of the cofactor is<br />
measured spectrophotometrically at 340 nm:<br />
C2H5OH ‡ NAD ‡ ˆ CH3CHO ‡ NADH ‡ H ‡<br />
CH3CHO ‡ NAD ‡ ‡ H2O ˆ CH3CO2H ‡ NADH ‡ H ‡<br />
As mentioned above, duty in the United Kingdom used to be levied on the gravity of<br />
the wort fermented. Accordingly the Institute of <strong>Brewing</strong> provide a method to determine<br />
the original gravity <strong>and</strong> Analytica-EBC <strong>and</strong> the ASBC give a method to find the original<br />
extract. Both are based on the distillation method for alcohol when, as well as the<br />
distillate, the residue in the distillation flask is diluted to the original volume <strong>and</strong> the<br />
specific gravity is measured. In the British method, the number of degrees of gravity by<br />
which the distillate is less than the specific gravity of distilled water is called the spirit<br />
indication of the distillate. From the `Mean Brewery Table' (Thorpe <strong>and</strong> Brown, 1914;<br />
HM Customs <strong>and</strong> Excise, 1997; Statutory Instrument No. 1146, 1979) the degrees of<br />
extract that must have been fermented to produce the spirit indication is read off <strong>and</strong><br />
added to the gravity of the residue to give the original gravity. Such a Table is necessary<br />
because in any fermentation carbohydrate is used for yeast growth <strong>and</strong> the production of<br />
metabolites other than ethanol, so the yield of ethanol is always less than that predicted<br />
by Gay-Lussac's equation:<br />
C6H12O6 ! 2C2H5OH ‡ 2 CO2<br />
Any radical change in the ratio of yeast growth to alcohol production from that used in<br />
compiling the `Mean Brewery Table' could result in beers in which the original gravity,<br />
determined by the above method, is higher than that actually employed. Thus, the use of<br />
larger fermentation vessels <strong>and</strong> methods of continuous fermentation results in beers in<br />
which the measured OG is higher than the gravity of the wort employed. Similarly, the<br />
production of large amounts of secondary metabolites can alter the results. Belgian<br />
lambic beer contains a high level of volatile acids <strong>and</strong> this is taken into account in the<br />
Belgian method for determining the original extract.<br />
In the EBC <strong>and</strong> ASBC methods for determining the original extracts of beers, the<br />
apparent extract (E % w/w) is determined from the specific gravity of the filtered beer,<br />
the alcohol content (A % w/w) from the specific gravity of the distillate, <strong>and</strong> the real<br />
extract (ER% w/w) is determined from the specific gravity of the distillation residue<br />
made up to the original volume. From Balling's equation when the extract in the original<br />
wort (% w/w) (ˆ ëPlato) ˆ p<br />
or<br />
p ˆ 100<br />
2:0665 A ‡ ER<br />
100 ‡ 1:0665 A<br />
p ˆ ER ‡ E<br />
‡ ER<br />
q<br />
where q is the correcting factor. The ASBC also define:<br />
Real degree of fermentation ˆ<br />
Apparent degree of fermentation ˆ<br />
100… p ER†<br />
p<br />
100… p E†<br />
p<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
For small breweries without laboratory facilities HM Customs <strong>and</strong> Excise (1997)<br />
provide an approximate formula to determine alcohol by volume:<br />
%ABV ˆ…OG PG† f<br />
whereOGistheoriginalgravityofthewort,PGisthepresentgravityofthebeer,<strong>and</strong>fis<br />
afactor connecting change in gravity with alcoholic strength. Unfortunately fis not<br />
constant <strong>and</strong> varies from 0.125 for weak beers to 0.135 for very strong beers. For the<br />
majority of popular UK beers fis 0.128±0.129. Brewers using this method should obtain<br />
confirmatory testing by an independent analyst at least annually.<br />
19.1.3 Carbohydrates<br />
The total carbohydrates remaining in beer can be estimated spectrophotometrically with<br />
anthrone in 85% sulphuric acid (Analytica-EBC, ASBC, IoB) for arange of beers values<br />
between 0.89±5.98% as glucose were found. Fully attenuated low carbohydrate `lite'<br />
beers, originally brewed for diabetic patients, are now generally available with<br />
carbohydrate contents of 0.4±0.9% w/v as glucose. Of the carbohydrates present in<br />
wort, glucose (4.1), fructose (4.2), sucrose (4.3), maltose (4.4) <strong>and</strong> maltotriose (4.5) will<br />
usually be fermented. Under-attenuating yeast strains will not ferment maltotriose while<br />
super-attenuating strains will partly ferment maltotetraose (4.6) but, in general, beers will<br />
contain only low levels of fermentable sugars other than those added as primings (Table<br />
19.5). Nevertheless trace amounts of many other sugars have been detected in beer<br />
including the monosaccharides; D-ribose (4.12), L-arabinose (4.11), D-xylose (4.10), Dmannose<br />
(4.8) <strong>and</strong> D-galactose (4.9), the disaccharides isomaltose (4.13), cellobiose<br />
(4.18) <strong>and</strong> kojibiose, <strong>and</strong> the trisaccharides panose (4.14) <strong>and</strong> isopanose (4.15). Data for<br />
these <strong>and</strong> other oligosaccharides are given in Table 19.6.<br />
Analytica-EBC <strong>and</strong> ASBC give methods for the fermentable carbohydrates in beer by<br />
HPLC. A detailed study of the dextrins in Tuborg lager beer was made using gel<br />
chromatography (Enevoldsen <strong>and</strong> Schmidt, 1974). The carbohydrates present were<br />
resolved into the following fractions (degree of polymerization-glucose units <strong>and</strong><br />
percentage of the total carbohydrates): DP 1±3, 7.4%; DP 4, 13.1%; Group I(DP 5±10),<br />
22.7%; GroupII(DP11±16),16.7%;GroupIII(DP17±21),9.7%;GroupIV(DP22±27),<br />
6.2%; Group V (DP 28±34), 4.0%; <strong>and</strong> higher dextrins (DP>35), 15.2%. Debranching<br />
studies with pullulanase indicated that the dextrins in Group Iwere either linear or single<br />
branched, while those in Groups II, III <strong>and</strong> IV contain two, three <strong>and</strong> four -(1±6)linkages<br />
respectively. The majority of the -(1±6)-linkages in amylopectin appear to<br />
survive the brewing process.<br />
If -(1±3)(1±4)-D-glucan, which makes up 70% of the barley endosperm cell wall<br />
(Section 4.4.3), is not completely broken down during malting, it may survive into beer<br />
where it can precipitate <strong>and</strong> lead to filtration problems. Only -glucans with molecular<br />
weights in excess of 200,000 daltons are said to precipitate. Two methods of analysis for<br />
-glucans are given in Analytica-EBC, one enzymatic using lichenase <strong>and</strong> the other using<br />
the fluorochrome Calcofluor. A range of beers contained 0±650 mg/l of -glucan; 10/36<br />
had no -glucan but a beer which precipitated a -glucan gel contained 1,900 mg/l. By<br />
colorimetric methods beers were found to contain 0.13±0.21 % of fructose <strong>and</strong> fructosans<br />
<strong>and</strong> 0.25±0.39 % of pentosans. Schwarz <strong>and</strong> Han (1995) found the arabinoxylan content<br />
of a number of beers was in the range of 514±4,211 mg/l.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.5 Sugar content of commercial beers shown as percentage (w/v) in sample (Otter <strong>and</strong> Taylor, 1967)<br />
Type of beer OG SG Fructose Glucose Sucrose Maltose Maltotriose Maltotetraose Total<br />
(hydrate)<br />
1 Pale ale 1050 1011 Nil 0.06 Nil 0.54 0.28 0.04 0.92<br />
2 Brown ale (primed) 1032 1012 1.0 1.0 Trace Trace 0.2 0.4 2.6<br />
3 Stout (primed) 1033 1013 0.53 0.61 Trace Trace 0.08 0.06 1.28<br />
4 Sweet stout (primed) 1045 1022 0.6 1.2 Trace Trace 0.6 0.3 3.6*<br />
5 Pale ale 1068 1019 0.01 0.01 Nil 0.7 1.7 0.4 2.9<br />
6 Strong ale 1085 1026 Trace Trace Nil 0.16 0.21 0.12 0.49<br />
7 Lager 1032 1007 Nil Nil Trace Trace Trace Trace Trace<br />
8 Lager (export) 1045 1008 Nil Trace Nil Trace 0.28 0.18 0.46<br />
9 Stout (conditioned) 1044 1008 Nil Nil Nil Nil Trace Nil Trace<br />
10 Ale 1038 1002 Trace 0.8 Nil Nil Trace Trace 0.8<br />
11 Lager 1040 1003 0.18 0.49 Nil Nil Nil Nil 0.67<br />
12 Lager 1046 1003 Nil Trace Nil Trace Trace Trace Trace<br />
13 Lager 1045 1004 Nil 0.27 Nil 0.17 0.24 0.09 0.77<br />
14 Lager 1052 1011 Trace 0.15 Nil 0.13 0.16 0.14 0.58<br />
15 Lager 1045 1008 Nil Nil Nil 0.25 0.33 0.20 0.78<br />
* Contained also lactose (hydrate) 0.9%<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.6 Oligosaccharides in beer (g/100ml as glucose)<br />
Lager Lager Diabetic<br />
(Danish)* (German)y lagery<br />
Pentose<br />
Fructose<br />
Glucose<br />
±<br />
} 0.02<br />
0.019<br />
0.015<br />
±<br />
0.052<br />
±<br />
0.008<br />
Isomaltose 0.08 0.102 0.098<br />
Maltose 0.07 0.188 0.143<br />
Panose ± 0.036 0.066<br />
Maltotriose 0.17 0.315 0.193<br />
4- -Isomaltosyl-D-maltose ± 0.049 0.100<br />
Maltotetraose 0.30 0.187 0.043<br />
Maltopentaose 0.08 0.144 0.100<br />
Maltohexaose 0.15 0.130 0.039<br />
Maltoheptaose 0.15 0.063 0.035<br />
Malto-octaose 0.17<br />
1.560 0.065<br />
Maltonoaose 0.15<br />
Higher dextrins 1.06 ± ±<br />
Total 2.40 2.830 0.092<br />
* Gjertsen (1955).<br />
y Silbereisen <strong>and</strong> Bielig (1961).<br />
} }<br />
19.1.4 Other constituents containing carbon, hydrogen <strong>and</strong> oxygen<br />
Non-volatile<br />
Manyofthesecomponentsareproductsofyeastmetabolism.Mostoftheintermediatesin<br />
the metabolic pathways discussed in Chapter 12 have been detected in beer.<br />
Quantitatively glycerol is important <strong>and</strong> arange of 436±3,971mg/l has been found; the<br />
highest level in aspecial beer of OE 27.3 g/100ml. In general, top fermented beers had<br />
higher glycerol levels than Pilsen-type beers (Klopper et al., 1986). Significant amounts<br />
of higher polyols have not been found but beer contains butane-2,3-diol (up to 280 mg/l)<br />
<strong>and</strong> smaller amounts of pentane-2,3-diol together with 3-hydroxybutan-2-one (9.16,<br />
acetoin, 3±26mg/l) <strong>and</strong> 3-hydroxypentan-2-one. These are reduction products of the<br />
volatile vicinal diketones (see later). Peppard <strong>and</strong> Halsey (1982) found 2,4,5-trimethyl-<br />
1,3-dioxolane (19.1) in beer (c. 0.1 mg/l); this is the cyclic acetal formed between<br />
butane-2,3-diol <strong>and</strong> ethanal (acetaldehyde). Similar 1,3-dioxolanes formed between<br />
butane-2,3-diol <strong>and</strong> isobutanal <strong>and</strong> isopentanal were also detected. Another non-volatile<br />
alcohol found in beer is tyrosol (19.2), the Ehrlich pathway degradation product of<br />
tyrosine. Canadian lager had 22.1±29.4mg/l while ales had 3.0±13.6mg/l, well below the<br />
taste threshold of 200ppm (McFarlane <strong>and</strong> Thompson, 1964).<br />
Non-volatile acids found in beer are listed in Table 19.7. In addition, Klopper et al.<br />
(1986) found pyruvic acid (1±127mg/l), malic acid (6±136mg/l), lactic acid (10±<br />
1362mg/l) <strong>and</strong> citric acid (6±211mg/l) in arange of beers; the highest levels of lactic<br />
acid(<strong>and</strong>aceticacid)werefoundinBelgian`acid'beers.Thelevelofoxalicacidinbeers<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
isimportant becausetheinsolublecalciumsaltmaycauseahaze<strong>and</strong>/orpromotegushing.<br />
Beers contain trace amount of lipids. ASwedish beer (12ëPlato) was found to contain<br />
(mg/l): triglycerides, 0.1±0.2; diglycerides, 0.1; monoglycerides, 0.1±0.3; sterol esters,<br />
0.01; free sterols, 0.01±0.02 <strong>and</strong> free fatty acids C 4±C 10, 10±15; <strong>and</strong> C 12±C 18, 0±0.5<br />
(AÈyraÈpaÈaÈet al., 1961). Similar data were found for other beers. The free fatty acids are<br />
volatile but the addition of another substituent usually results in loss of volatility.<br />
Autoxidation of linoleic acid gives rise to isomers of dihydroxy- <strong>and</strong> trihydroxyoctadecenoic<br />
acids. The concentration of these acids (Table 19.7) is greater than that of<br />
linoleic acid itself. These hydroxy acids are potential precursors of 2-trans-nonenal,<br />
which contributes acardboard flavour to stale beer. The influence of lipids on foam <strong>and</strong><br />
head retention is discussed later.<br />
Also included in Table 19.7 are the phenolic acids present in beer which, with other<br />
polyphenols, are extracted from malt <strong>and</strong> hops (Sections 4.8 <strong>and</strong> 8.4). Polyphenols give<br />
colours with ferric salts <strong>and</strong> Analytica-EBC <strong>and</strong> the ASBC give a non-specific<br />
specrophotometric method for polyphenols based on the colour formed at 600nm with<br />
ferric ammonium citrate. (‡)-Catechin (4.138) can be used as st<strong>and</strong>ard. In addition<br />
Analytica-EBC give aspectrophotometric method for flavanoids based on the colour<br />
formed with p-dimethylaminocinnamaldehyde: again (‡)-catechin may be used as<br />
st<strong>and</strong>ard. By HPLC the polyphenols quercetin (8.58, RˆH)(36±148ppm), rutin (8.58,<br />
R ˆ -L-rhamnosyl-6- -D-glucosyl) (1.5±7.7ppm), catechin (26Ð141ppm) <strong>and</strong><br />
epicatechin (4.139) (8.5±127ppm) have been quantified in beer (Qureshi et al., 1979).<br />
Beers also contain proanthocyanidins (anthocyanogens) of which the dimeric<br />
procyanidin B-3 (4.143) predominates (0.5±4.0ppm). Whittle et al. (1999) examined<br />
the polyphenols in sixty beers by HPLC with electrochemical detection. They found over<br />
twenty procyanidin dimers <strong>and</strong> trimers in beer but no tetramers or pentamers although<br />
thesewerepresentinbarleyextracts.Theyalsoobservedwhichpeakswereremovedwhen<br />
the beer was treated with excess polyvinylpyrrolidone (20g/l). As the name implies,<br />
proanthocyanidinsontreatmentwithacidformthecolouredanthocyanidinpigments( max<br />
c. 545 nm) but the reaction is not straightforward <strong>and</strong> the yield of pigment from different<br />
products varies. Polyphenols, particularly proanthocyanidins, react with proteins during<br />
the storage of beer to produce non-biological haze (see later) but the yields of<br />
anthocyanidins, liberated with acid, do not correlate with the shelf-life of the beer. The<br />
haze potential of beers has been estimated nephelometrically by the haze formed after<br />
treatment with either cinchonidine sulphate, polyvinylpyrrolidone 700 or tannic acid.<br />
CloselyrelatedtothepolyphenolsarethehopresinsdiscussedinChapter8.Themajor<br />
bittering principles in beer are the cis- <strong>and</strong> trans-isomers of isocohumulone, isohumulone<br />
<strong>and</strong> isoadhumulone (8.40). The individual isomers may be resolved by HPLC but for<br />
routine analysis they are usually estimated together from the light absorption of an isooctane<br />
extract of acidified beer at 275nm (Analytica-EBC, IoB <strong>and</strong> ASBC). To avoid<br />
making assumptions about the nature of the bittering principles, the absorbance is<br />
multiplied by 50 <strong>and</strong> the result given as (International) Bitterness Units (IBU or BU).<br />
Nevertheless, in beers brewed with fresh hops IBU approximate to mg iso- -acids/l.<br />
Beers maycontain 10±60IBU(Tables19.1<strong>and</strong>19.2) <strong>and</strong>exceptionally 100IBU(Glaser,<br />
2002). The sensory detection threshold of the iso- -acids is c. 5±6 mg/l <strong>and</strong> the<br />
tetrahydro- <strong>and</strong> hexahydroiso- -acids are even more bitter (Weiss et al., 2002).<br />
Volatile<br />
Although trace amounts of the volatile constituents of malt <strong>and</strong> hops may survive wort<br />
boiling, the majority of the volatile constituents of beer are fermentation products. After<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.7 Non-volatile acids in beer<br />
Acid (ppm) Berlin Pilsner (a) German (b) British (c) American (d) Flavour threshold (e)<br />
C2 Glycolic ± ± ± ±<br />
Oxalic (4.151) ± 9.9±22.8 ± 0.5±3.0<br />
C3 D-Lactic ± 20±200 ± ±<br />
L-Lactic ± 40±152 ± ±<br />
Lactic (4.150) 188 ± 44±292 ± (400)<br />
Pyruvic (4.146) ± 42±75 10±104 ± (300)<br />
Malonic 0.02 ± ± ±<br />
C4 Succinic (4.149) 48 ± 36-166 ±<br />
Fumaric (4.148) ± ± ± ±<br />
Malic (4.152) ± 55±105 14±97 ±<br />
Oxaloacetic ± ± ± ± (500)<br />
Tartaric ± ± ± ± (600)<br />
C 5 2-Hydroxy-3-methylbutyric 0.26 ± ± ±<br />
Levulinic ± ± ± ±<br />
2-Methylfumaric (Mesaconic) ± ± ± ±<br />
Glutaric 0.01 ± ± ±<br />
2-Hydroxyglutaric ± ± 0±17 ±<br />
Citramalic ± ± ± (5.9±15.2)<br />
2-Oxoglutaric (4.147) ± ± 0±20<br />
C6 2-Hydroxy-3-methylpentanoic 0.29 ± ± ±<br />
2-Hydroxy-4-methylpentanoic 0.33 ± ± ±<br />
Adipic ± ± ± ±<br />
Kojic ± ± ± 5±78.2<br />
Citric (4.153) ± 130±230 56±158 ±<br />
Isocitric ± ± ± ±<br />
Oxalosuccinic ± ± ±<br />
C7 Benzoic 0.45 ± ± ±<br />
2-Hydroxybenzoic 0.02 ± ± 1.00±9.0<br />
4-Hydroxybenzoic 0.13 ± ± ±<br />
2-Hydroxyheptanoic 0.06 ± ± ±<br />
3,4-Dihydroxybenzoic 2.4 ± ± 6.3±29.0<br />
2,5-Dihydroxybenzoic (Gentisic) ± ± ± 2.8±12.7<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Pimelic 0.01 ± ± ±<br />
Gallic (4.125) ± ± ± 12.0±30 360<br />
C 8 Phenylacetic 0.93 ± ± ± 2.5<br />
2-Hydroxyoctanoic 0.04 ± ± ±<br />
3-Hydroxyoctanoic 0.07 ± ± ±<br />
2-Ethylhexanoic ± ± ± 10.2 11±32<br />
4-Hydroxyphenylacetic 0.04 ± ± ±<br />
Suberic 0.15 ± ± ±<br />
Vanillic 2.4 ± ± 0.3±1.5 80<br />
Phthalic 0.02 ± ± ±<br />
C9 Phenylpropionic 0.01 ± ± ±<br />
trans-Cinnamic 0.5 ± ± 1.0±8.3<br />
cis-Cinnamic < 0.01 ± ± (trans + cis)<br />
Phenyl-lactic 1.2 ± ±<br />
4-Hydroxyphenylpropionic 0.02 ± ± ±<br />
trans-p-Coumaric 1.9 ± ± 8.21 520<br />
cis-p-Coumaric 0.02 ± ± (trans + cis)<br />
Caffeic ± ± ± 1.4±8.0 690<br />
Azeleic 1.5 ± ± ±<br />
C10 3-Hydroxydecanoic 0.16 ± ± ±<br />
trans-Ferulic 4.6 ± ± 1.7±20.8 660<br />
cis-Ferulic 1.1 ± ± (trans + cis)<br />
C11 Undecanedioic 0.13 ± ± ±<br />
Sinapic ± ± ± 0.7±3.6<br />
C12 Dodecanedioic 0.18 ± ±<br />
C16 Chlorogenic ± ± ± 1±11.2<br />
C18 Trihydroxyoctadecanoic<br />
9,12,13-10-trans ± 4.9±9.0 ±<br />
9,12,13-11-trans ± 1.0±2.4 ±<br />
9,10,11-12-trans ± 0.4±0.7 ±<br />
[a] Tressl et al. (1975)<br />
[b] M<strong>and</strong>l <strong>and</strong> Piendl (1971)<br />
[c] Coote <strong>and</strong> Kirsop (1974)<br />
[d] Qureshi et al. (1979)<br />
[e] Meilgaard (1975)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.8 Volatile constituents of beer<br />
(a) Alcohols [a]<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C 1 Methanol ± 10,000<br />
C 2 Ethanol ± 14,000<br />
C 3 Propanol ± 800<br />
Propan-2-ol ± 1,500<br />
C4 Butanol ± 450<br />
Butan-2-ol ± 16<br />
2-Methylpropanol 4.8 200<br />
C5 Pentanol 0.15 (80)<br />
Pentan-2-ol ± 45<br />
2-Methylbutanol 84.0 70<br />
3-Methylbutanol (2 Me +3Me) 65<br />
Furufuryl alcohol 1.20 3,000<br />
C6 Hexanol 0.33 4.0<br />
Hexan-2-ol ± 4.0<br />
Hex-2-enol 0.025 13<br />
Hex-3-enol 0.020 15<br />
C7 Heptanol ± 1.0<br />
Heptan-2-ol 0.015 0.25<br />
Benzyl alcohol ± 900<br />
C 8 Octanol ± 0.9<br />
Octan-2-ol 0.005 0.04<br />
Oct-l-en-3-ol 0.030 0.2<br />
2-Phenylethanol 1.8 125<br />
Tyrosol see text 200<br />
4-Vinylphenol 0.025 ±<br />
C9 Nonanol ± 0.08<br />
Nonan-2-ol 0.01 0.075<br />
4-Vinylguaiacol 0.10 0.3<br />
C 10 Decanol ± 0.18<br />
Decan-2-ol 0.005 0.015<br />
Linalol ± 0.08<br />
-Terpinol ± 2.0<br />
Nerol ± 0.50<br />
C 12 Dodecanol ± 0.40<br />
Dodecan-2-ol ± ±<br />
(b) Aldehydes [a]<br />
Concentration Flavour threshold<br />
(ppm) [d] (ppm) [c]<br />
C2 Acetaldehyde See Table 19.10 10<br />
Glyoxal ± ±<br />
C3 Propanal ± 30<br />
Prop-2-enal (acrolein) 1.6 15<br />
Pyruvaldehyde ± -<br />
C 4 Butanal 10.9 (1.0)<br />
2-Methylpropanal ± (1.0)<br />
But-2-enal (crotonal) 1.33 8<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.8 Continued<br />
Concentration Flavour threshold<br />
(ppm) [d] (ppm) [c]<br />
C5 Pentanal 2.7 0.5<br />
2-Methylbutanal ± 1.25<br />
3-Methylbutanal 7.0 0.6<br />
Pent-2-enal 0.59 ±<br />
Furfural (25.0) 150<br />
C6 Hexanal 1.6 0.3<br />
Hex-2-enal 0.36 0.5<br />
5-Hydroxymethylfurfural see text 1,000<br />
C7 Heptanal 1.2 0.05<br />
Hept-2-enal 0.08 0.0005<br />
Benzaldehyde 10 2<br />
C8 Octanal 1.4 0.04<br />
Oct-2-enal 0.03 0.0003<br />
2-Phenylacetaldehyde (5) 1.6<br />
C9 Nonanal 3.7 0.015<br />
Non-2-enal 0.07 0.0003<br />
C 10 Decanal 0.9 0.005<br />
Dec-2-enal trace 0.001<br />
C11 Undecanal 0.4 0.002<br />
C12 Dodecanal 0.2 0.002<br />
(c) Acids [a]<br />
Concentration Flavour threshold<br />
(ppm) [e] (ppm) [c]<br />
C1 Formic ± ±<br />
C2 Acetic 175<br />
C3 Propionic 150<br />
C4 Butyric 0.62 2.2<br />
2-Methylpropionic 1.1 30<br />
Crotonic ± ±<br />
C5 Pentanoic (Valeric) 0.03 8<br />
2-Methylbutyric ± ±<br />
3-Methylbutyric 1.3 1.5<br />
Pentenoic ± ±<br />
C6 Hexanoic (Caproic)<br />
Hex-2-enoic<br />
2.5<br />
}<br />
8<br />
0.01<br />
Hex-3-enoic<br />
±<br />
1.3<br />
4-Methylpent-3-enoic 0.32 ±<br />
3-Carbethoxypropionic 0.22 ±<br />
C7 Heptanoic (Oenanthic) 0.03 ±<br />
Hept-2-enoic < 0.01 ±<br />
4-Methylhex-2-enoic ± ±<br />
Benzoic 0.45 ±<br />
C8 Octanoic (Caprylic) 6.1 13/15<br />
6-Methylheptanoic ± ±<br />
Oct-2-enoic < 0.01 ±<br />
Phenylacetic 0.93 2.5<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.8 Continued<br />
Concentration Flavour threshold<br />
(ppm) [e] (ppm) [c]<br />
C9 Nonanoic (Pelargonic) 0.02<br />
Non-2-enoc
Table 19.8 Continued<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C8 Methyl heptenoate ± ±<br />
Ethyl hexanoate 0.95 0.23<br />
3-Methylbutyl propionate 0.15 0.7<br />
Hexyl acetate 0.025 1.4<br />
Methyl 4-methylhex-2-enoate 0.06 ±<br />
Ethyl hexenoate ± ±<br />
Ethyl nicotinate 1.4 ±<br />
C9 Methyl octanoate ± ±<br />
Ethyl heptanoate ± ±<br />
Amyl butyrate ± 0.6<br />
3-Methylbutyl butyrate ± ±<br />
3-Methylbutyl 2-methylpropionate 0.140 ±<br />
Heptyl acetate 0.025 1.4<br />
EthyI heptenoate ± ±<br />
Ethyl benzoate 0.01 ±<br />
C10 Ethyl octanoate 1.50 0.9<br />
Butyl hexanoate ± ±<br />
2-Methylpropyl hexanoate 0.025 ±<br />
Amyl 3-methylbutyrate ± ±<br />
3-Methylbutyl 3-methlbutyrate 0.040 ±<br />
Hexyl butyrate<br />
Octyl acetate 0.030 0.5<br />
2-Phenylethyl acetate 1.625 3.8<br />
C11 Ethyl nonanoate ± 1.2<br />
Amyl hexanoate ± ±<br />
Pent-2-yl hexanoate ± ±<br />
3-Methylbutyl hexanoate 0.420 0.9<br />
Heptan-2-yl butyrate ± ±<br />
Nonyl acetate 0.005 ±<br />
Ethyl nonenoate 0.045 ±<br />
2-Phenylethyl propionate 0.010 ±<br />
Ethyl cinnamate 0.005 ±<br />
C12 Ethyl decanoate 0.190 1.5<br />
Butyl octanoate ± ±<br />
3-Methylbutyl heptanoate 0.020 ±<br />
Hexyl hexanoate ± ±<br />
Octyl butyrate ± ±<br />
2-Phenylethyl butyrate ± ±<br />
3-Phenylethyl 2-methypropionate 0.015 ±<br />
3-Methylbutyl benzoate ± ±<br />
Ethyl dec-4-enoate 0.035 ±<br />
Ethyl deca-4, 8-dienoate 0.015 ±<br />
C13 2-Methylbutyl octanoate 0.015 ±<br />
3-Methylbutyl octanoate 0.830 2.0<br />
2-Phenylethyl 3-methylbutyrate 0.015 ±<br />
C 14 Ethyl dodecanoate 0.030 3.5<br />
3-Methylbutyl nonanoate ± 2.0<br />
Hexyl octanoate ± ±<br />
Octyl hexanoate ± ±<br />
2-Phenylethyl hexanoate 0.075 ±<br />
2-Phenylethyl hexenoate ± ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.8 Continued<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C15 2-Methylbutyl decanoate 0.005 ±<br />
3-Methylbutyl decanoate 0.095 3.0<br />
3-Methylbutyl dec-4-enoate 0.020 ±<br />
C 16 Ethyl tetradecanoate ± 2.5<br />
Hexyl decanoate ± ±<br />
2-Phenylethyl octanoate 0.015 ±<br />
C 17 3-Methylbutyl dodecanoate 0.010 ±<br />
(e) Lactones [a]<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C 4 4-Butanolide 1.300 ±<br />
C6 4-Hexanolide 0.0200<br />
4, 4-Dimethylbutan-4-olide 0.160 ±<br />
4, 4-Dimethylbut-2-en-4-olide 1.750 ±<br />
C7 4-Heptanolide 0.015 ±<br />
C8 4-Octanolide 0.020 ±<br />
C 9 4-Nonanolide 0.320 ±<br />
C10 4-Decanolide 0.020 0.4<br />
C 11 4-Dihydroactindiolide 0.030 ±<br />
(f) Ketones [a]<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C 4 Butan-2-one 0.018 (80)<br />
Butane-2, 3-dione 0.058 0.15<br />
3-Hydroxybutan-2-one 0.420 (50)<br />
C5 Pentan-2-one 0.020 (30)<br />
Pentan-3-one ± (30)<br />
Pentane-2, 3-dione 0.012 0.9<br />
3-Hydroxypentan-2-one 0.050 ±<br />
2-Methyltetrahydrofuran-3-one 0.025 ±<br />
2-Methyltetrahydro-thiophen-3-one 0.005 ±<br />
C6 Hexan-2-one ± 4.0<br />
3-Methylpentan-2-one 0.060 0.4<br />
4-Methylpentan-2-one 0.12 ±<br />
2-Acetylfuran 0.040 ±<br />
C 7 Heptan-2-one 0.110 2.0<br />
C 8 Octan-2-one 0.010 (0.25)<br />
6-Methylhept-5-en-2-one 0.050<br />
C 9 Nonan-2-one 0.030 (0.2)<br />
C10 Decan-2-one ± 0.25<br />
C11 Undecan-2-one 0.001 0.4<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.8 Continued<br />
(g) Hydrocarbons [a]<br />
Concentration Flavour threshold<br />
(ppm) [b] (ppm) [c]<br />
C6 Cyclohexane ± ±<br />
C7 3, 4-Dimethylpent-2-ene ± ±<br />
Toluene ± ±<br />
C8 2, 2-Dimethylhexane ± ±<br />
o-Xylene 0.020 ±<br />
Styrene 0.070 ±<br />
C 9 Methylethylbenzene 0.020 ±<br />
C 10 i-Butylbenzene ± ±<br />
-Pinene ± ±<br />
Limonene ± ±<br />
Myrcene ± 0.013<br />
Naphthalene 0.015 ±<br />
C15 Carophyllene ± ±<br />
[a] Drawert <strong>and</strong> Tressl (1972)<br />
[b] Tressl et al. (1978)<br />
[c] Meilgaard (1975)<br />
[d] Greenhoff <strong>and</strong> Wheeler (1981)<br />
[e] Tressl et al. (1975)<br />
ethanol, discussed above, the largest group of volatile constituents are the higher<br />
alcohols; those identified in beer are listed in Table 19.8(a) By distillation the higher<br />
alcohol fraction may be separated when it is known as fusel oil. Thus, the distiller has a<br />
closercontroloverthehigheralcoholcontentofhisbeveragethanthebrewer.Gin,vodka<br />
<strong>and</strong> grain whisky have low levels of higher alcohols while malt whisky <strong>and</strong> br<strong>and</strong>y,<br />
produced in pot stills, usually have higher levels of these congeners.<br />
The principal higher alcohols found in beer are 3-methylbutanol (isoamyl alcohol), 2methylbutanol<br />
(active-amyl alcohol), 2-methylpropanol (isobutyl alcohol), propanol,<br />
(propyl alcohol) <strong>and</strong> -phenylethanol (phenethyl alcohol) (Table 19.9). Greenshields<br />
(1974) found that the level of higher alcohols in home-brewed beers <strong>and</strong> wines was ten<br />
times higher than the level in commercial products. The major volatile constituents of<br />
beer are most conveniently examined by gas chromatography. The results given in Table<br />
19.9 (Morgan, 1965) were obtained by GC after distillation <strong>and</strong> extraction of the volatile<br />
productsintoether.DirectinjectionofbeerontoasuitableGCcolumnminimizessample<br />
preparation but requires frequent replacement of the top layer of apacked column or of<br />
the glass wool in the injection port due to the deposition of beer solids. Although the<br />
precision is not high (Baker, 1989) aheadspace method of analysis of the major volatiles<br />
in beer has been approved giving values for acetaldehyde, propanol, isobutanol,<br />
methylbutanols, ethyl acetate <strong>and</strong> ethyl hexanoate (Fig. 19.1).<br />
In order to identify the minor volatile constituents of beer it is usually necessary to<br />
examine adistillate or solvent extract which can be fractionated further by adsorption<br />
chromatography (Tressl et al., 1975, 1978, Lermusieau et al., 2001).<br />
Also included in Table 19.8(a) are 4-vinylphenol <strong>and</strong> 4-vinylguaicol (4.134, 4hydroxy-3-methoxystyrene),<br />
which are regarded as off-flavours in most beers. However,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.9 Principal volatile constituents of beer (Morgan, 1965)<br />
Concentration Range<br />
B.p. ( o C) Stout Pale ale Brown ale Lager<br />
Ethanol (% v/v) 78 2.0±8.9 3.4±4.0 2.1±4.5 2.8±3.2<br />
(mg/l)<br />
n-Propanol 97 13±60 31±48 17±29 5-10<br />
2-Methylpropanol 108 11±98 18±33 11±33 6±11<br />
2-Methylbutanol 128 9±41 14±19 8±22 8±16<br />
3-Methylbutanol 131 33±169 47±61 28±77 32±57<br />
-Phenylethanol 220 20±55 36±53 19±44 25±32<br />
Ethyl acetate 77 11±69 14±23 9±18 8±14<br />
Isoamyl acetate 139 1.0±4.9 1.4±3.3 0.4±2.6 1.5±2.0<br />
4-vinylguaicol, which has aclove-like, spice-like flavour, provides part of the essential<br />
character of Weizenbier <strong>and</strong> Rauchbier. Arange of ale <strong>and</strong> lagers contained 0±0.09mg/l<br />
4-vinylguaicol, below the threshold value of 0.3mg/l, but Weissbier contains 0.8±<br />
1.5mg/l (McMurrough et al., 1996). These phenols are formed by decarboxylation of<br />
trans-p-coumaric acid (4.129) <strong>and</strong> trans-ferulic acid (4.131) respectively. This may<br />
occur thermally during kilning or wort boiling or enzymatically during fermentation.<br />
The capacity of yeasts to decarboxylate cinnamic acids (Pof‡ phenotype) is strong in<br />
wild strains of Saccharomyces but absent from lager-brewing yeasts <strong>and</strong> most alebrewing<br />
yeasts. However, it is present in the yeasts used in the production of wheat<br />
beers.Thelevelof4-vinylguaicolinbeerdeclinesduringstorage(withahalflifeofc.60<br />
days at 18ëC) presumably due to the formation of 4-(1-ethoxyethyl)guaiacol.<br />
Decarboxylation of cinnamic acid will give styrene (Table 19.8(g)), which is reported<br />
tobecarcinogenic.AccordinglyPof‡strainsofyeastareavoidedinbrewingmostbeers.<br />
Onlylowlevelsofaldehydesarefoundinbeer(Table19.8(b)).AsdiscussedinChapter<br />
12, ethanol <strong>and</strong> the higher alcohols are formed by reduction of the corresponding<br />
aldehydes by the enzyme alcohol dehydrogenase. Acetaldehyde (ethanal) is the major<br />
aldehyde in beer <strong>and</strong> some values are given in Table 19.10. Acetal (1,1-diethoxyethane),<br />
40<br />
1<br />
2 3<br />
4<br />
5<br />
7<br />
6<br />
8<br />
0 0 16<br />
10<br />
9<br />
11<br />
12<br />
13<br />
14<br />
GRAPH KEY<br />
40 mV %<br />
16.00 min.<br />
1. Acetaldehyde<br />
2. Dimethyl sulphide<br />
3. Acetone<br />
4. Ethyl acetate<br />
5. Methanol<br />
6. Isobutyl acetate<br />
7. Ethyl butyrate<br />
8. n-Propanol<br />
9. Isobutanol<br />
10. Isoamyl acetate<br />
11. Internal st<strong>and</strong>ard<br />
12. Isopentanols<br />
13. Ethyl hexanoate<br />
14. Ethyl octanoate<br />
Fig. 19.1 Typical chromatogram of volatile compounds in beer (Institute of <strong>Brewing</strong>, Methods<br />
of Analysis, 1997)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.10 Acetaldehyde content of beers (Otter <strong>and</strong> Taylor, 1971)<br />
Beer Acetaldehyde content (ppm)<br />
(mean value in parentheses)<br />
British lager 2.3±28.2 (8.7)<br />
Foreign lager 0±13 (5.4)<br />
Light <strong>and</strong> pale ales 3.8±33.8 (8.2)<br />
Primed beers 3.0±37.2 (15.2)<br />
Irish stout 0.5±10.0 (4.0)<br />
American beers 2±18 (9.1)<br />
formed by condensation of ethanal with two molecules of ethanol, has been detected in<br />
beer. During the storage of bottled beer higher alcohols are oxidized to aldehydes by<br />
melanoidins; these aldehydes have much lower threshold values than the parent alcohols<br />
<strong>and</strong> can produce off-flavours. As mentioned above, the cardboard flavour of stale beer is<br />
though to be due to 2-trans-nonenal <strong>and</strong> 5-methylfurfural (9.22). Among the heterocyclic<br />
compounds formed during wort boiling (Chapter 9), 5-hydroxymethylfurfural (9.8, R=<br />
CH2OH), 5-methylfurfural <strong>and</strong> furfural (9.17) itself are found in beer. One survey of over<br />
300 German beers reported 0.5±4 ppm of 5-hydroxymethylfurfural (maximum value 7.8<br />
ppm) (Thalacker <strong>and</strong> Kaltwasser, 1978) but other workers found higher levels increasing<br />
with beer colour; one dark German beer contained 71.5 ppm (Kieninger <strong>and</strong> Birkova,<br />
1975). Lower levels of furfural are found (
ethyl acetate <strong>and</strong> 107±483ppm of ethyl lactate (Van Oevelen et al., 1976). The majority<br />
of the esters in beer are fermentation products but traces of esters from hop oil may be<br />
present. Methyl 4-decenoate <strong>and</strong> methyl 4,8-decadienoate, present in hop oil, are<br />
transesterified during fermentation to give the corresponding ethyl esters in beer. No<br />
doubt other esters in wort are transesterified during fermentation to give ethyl esters in<br />
beer.Lactones(Table19.8(e))arecyclicestersofhydroxyacids.Dihydro-2(3H)-furanone<br />
(19.3) <strong>and</strong> dihydro-5-methyl-2(3H)-furanone may be regarded as lactones.<br />
Ketones (Table 19.8(f)) are, like aldehydes, carbonyl compounds but they are not<br />
major fermentation products. Many of those present in beer may be derived from hop oil<br />
or hop resin degradation products. An exception may be 2-acetylfuran which occurs at<br />
higher levels in ales (90±97ppm) than in lagers (4±12ppm). 4-Hydroxy-3(2H)-furanone<br />
(19.4) derivatives also contribute to beer flavour. 5-Methyl-4-hydroxy-3(2H)-furanone<br />
has ameaty <strong>and</strong> brothy flavour note but the level in beer (0.12±0.77ppm) is less than the<br />
flavour threshold (8.3ppm). 2,5-Dimethyl-4-hydroxy-3(2H)-furanone has a sweet<br />
caramel flavour (threshold 0.16ppm) so the concentration in beer (0.19±2.73ppm) will<br />
cause it to influence flavour. 2-(or 5)-Ethyl-5-(or 2)-methyl-4-hydroxy-3(2H)-furanone<br />
also has asweet caramel flavour with alower threshold (0.02 ppm) than the dimethyl<br />
compound. It has been detected in some beers but not in others; in all cases the<br />
concentration was less than the threshold level (Mackie <strong>and</strong> Slaughter, 2000).<br />
Vicinal or -diketones have two carbonyl groups on adjacent carbon atoms. The most<br />
important vicinal diketone (VDK) in beer is diacetyl (9.14, butane-2,3-dione) which is<br />
accompanied by smaller amounts of pentane-2,3-dione. These diketones produce<br />
butterscotch flavours with thresholds of 0.07±0.15mg/l <strong>and</strong> 0.9mg/l respectively.<br />
Quantities in excess of 0.15mg/l of diacetyl are said to produce an off-flavour in lager<br />
beer but higher levels may be acceptable in ales <strong>and</strong> stouts. Many methods have been<br />
proposed for the analysis of diacetyl <strong>and</strong> vicinal diketones. The most specific involves<br />
gas liquid chromatography with an electron capture detector which is sensitive to vicinal<br />
diketones but not tothe majority ofotherbeer constituents (Harrison et al., 1965a,b). The<br />
Institute of <strong>Brewing</strong> has described methods using both packed <strong>and</strong> capillary columns<br />
(Buckee <strong>and</strong> Mundy, 1994) <strong>and</strong> the latter, which uses hexane-2,3-dione as internal<br />
st<strong>and</strong>ard, has been accepted by the EBC. The analyses given in Table 19.11 were<br />
obtained using such methods. In addition there are several colorimetric methods for<br />
determining vicinal diketones. The vicinal diketones are by-products of the biosynthesis<br />
of the amino acids valine <strong>and</strong> leucine (Fig. 12.21). One of the intermediates, -<br />
acetolactate, decomposes to diacetyl on heating so analytical methods which involve a<br />
preliminary distillation, especially those of fermenting wort, give high results. Headspace<br />
analysis avoids distillation <strong>and</strong> using this technique, with electron capture detection, it<br />
was found that during fermentation of an Irish stout the maximum level of vicinal<br />
diketones (0.6 ppm) was found 44 h after pitching but thereafter the level fell to about<br />
0.1 ppm. Yeast strains differ in the amount of diacetyl they produce <strong>and</strong> the choice of<br />
yeast strain is probably the most important factor in controlling the level of VDK in beer<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.11 Diacetyl <strong>and</strong> Pentane-2,3-dione content of beers (mg/l) (Harrison et al., 1965a,b)<br />
Quality* Diacetyl Pentane-2,3-dione<br />
Barley wine (4) 0.11±0.40 0.04±0.08<br />
Lager (9) 0.02±0.08 0.01±0.05<br />
Ale (9) 0.06±0.30 0.01±0.20<br />
Stout (5) 0.02±0.07 0.01±0.02<br />
Stout (1) 0.58 0.26<br />
* Number of samples in parentheses.<br />
(Portno, 1966). In particular respiratory deficient `petite mutants' of yeast produce large<br />
quantities of diacetyl <strong>and</strong> the off-flavour produced in beers infected with Pediococci (the<br />
so-called beer sarcina) is due to diacetyl. During fermentation ketones may be reduced to<br />
secondaryalcohols.Thusreductionofdiacetylgivesfirst3-hydroxybutan-2-one(acetoin)<br />
<strong>and</strong> then butane-2,3-diol which are much less volatile. They are also much less potent<br />
flavouring agents with threshold values of 17 <strong>and</strong> 4,500mg/l respectively. The addition<br />
of actively fermenting wort to beer may reduce the diacetyl content. The hop oil<br />
constituents which have been detected in beer are discussed in Chapter 8.<br />
19.1.5 Nitrogenous constituents<br />
Non-volatile<br />
During the 20th century the total nitrogen content of barley, malt <strong>and</strong> beer was usually<br />
measured by the Kjeldahl method where the sample is digested with concentrated<br />
sulphuricacid <strong>and</strong> asuitablecatalyst <strong>and</strong>the nitrogenous constituentsare brokendown to<br />
ammonium sulphate. After dilution of the digest, the ammonia is liberated <strong>and</strong> distilled<br />
intost<strong>and</strong>ardacid.AlthoughtheKjeldahlmethodhasbeenautomateditstillemploystoxic<br />
<strong>and</strong> hazardous reagents <strong>and</strong> today is being replaced by the older, but now automated,<br />
Dumas combustion method (Buckee, 1994, 1995, 1997; Johnson <strong>and</strong> Johansson, 1999).<br />
Herethesampleiscombustedinthepresenceofoxygenatabout1,000ëCtogiveoxidesof<br />
nitrogenwhicharecatalyticallyreducedtonitrogen.Otherproductsofcombustionsuchas<br />
carbondioxide <strong>and</strong>water areremovedby selective absorption<strong>and</strong>theremaining nitrogen<br />
measured in athermal conductivity cell. The Dumas method consistently gives slightly<br />
higher total nitrogen values than those given by the Kjeldahl method but it long been<br />
known that the Kjeldahl method does not deal efficiently with certain types of compound<br />
(Buckee,1995).Free -aminonitrogenisdeterminedbyaninternationallyagreedmethod<br />
measuring the purple colour formed with ninhydrin at 570nm.<br />
The total nitrogen content multiplied by the factor 6.25 is often expressed as `protein'.<br />
Mostbeerscontain 300±1,000mg/ltotal-Nequivalent to0.11±0.63% protein.Anall-malt<br />
Burton strong ale contained 1,840mg/l total-N equivalent to 1.15% protein. There<br />
appears to be no universally accepted definition of protein. Some authorities classify<br />
proteinsbyfunction,othersbytheirmolecularsize.Itisagreedthathydrolysisofproteins<br />
gives polypeptides, peptides <strong>and</strong> eventually amino acids (terms such as `proteoses <strong>and</strong><br />
`proteids' are even less well defined). One text suggests that peptides may contain up to<br />
10 amino acid residues, polypeptides 11±100 residues <strong>and</strong> proteins more than 100<br />
residues but Bamforth (1985) suggests that the term protein should be restricted to an<br />
undegraded molecule with an independent <strong>and</strong> unique identity such as, for example, a<br />
moleculeofanenzyme.Onthisbasishesuggestedthatfewproteinssurviveintobeer<strong>and</strong><br />
most of the nitrogen is present as polypeptides. Williams et al. (1995) compared seven<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
methods for determining the protein concentration of beer including both classical <strong>and</strong><br />
automated Kjeldahl analyses. They concluded that the protein dye-binding assays<br />
(Coomassie Brilliant Blue <strong>and</strong> Pyrogallol Red-Molybdate methods) gave reproducible<br />
values for the protein concentration consistent with the results indicated by electrophoresis.<br />
Many new techniques have been used to study the proteins/polypeptides present in<br />
beerprincipallywiththeaimofcharacterizingthefractionsassociatedwithfoamstability<br />
<strong>and</strong> haze formation. Dialysis with Visking tubing retains compounds with molecular<br />
weight greater than 5,000 which accounts for approximately half of the total, N. Size<br />
exclusion chromatography (gel filtration) of beer shows the presence of polypeptides<br />
across the range of molecular weights from 2,000 to >100,000 but the majority have Mr<br />
Table 19.12 Amino acid analyses of beers (amino acids as g -amino N/ml)<br />
Indian pale ale Brown ale Draught bitter Mild ale Stout Stout<br />
OG 1042.8 1031.5 1040.8 1036 1045 1045<br />
Alanine (4.24) ± 0.19 0.43 ± 0.2 ±<br />
Ammonia 1.07 0.92 1.88 1.36 1.7 1.8<br />
Arginine (4.28) ± ± 0.18 ± ± ±<br />
Aspartic acid (4.29) ± 0.04 0.16 0.4 0.2 0.2<br />
Glutamic acid (4.33) ± 0.04 0.14 0.2 ± ±<br />
Glycine (4.35) 0.21 0.01 0.08 0.19 ± ±<br />
Histidine (4.36) 0.14 ± ± ± ± ±<br />
Isoleucine (4.38) ± ± 0.01 ± 0.4 ±<br />
Leucine (4.39) ± ± 0.03 ± 0.4 ±<br />
Lysine (4.40) 0.42 0.31 0.50 0.32 0.2 0.6<br />
Methionine (4.41) ± ± 0.24 ± ± ±<br />
Phenylalanine (4.42) ± 0.04 0.49 0.07 ± ±<br />
Serine (4.45) ± ± 0.15 0.05 0.2 0.3<br />
Threonine (4.46) ± ± 0.66 ± 0.2 0.2<br />
Tryptophan (4.47) 1.77 1.62 2.06 1.62 0.5 1.2<br />
Tyrosine (4.48) ± 0.50 0.35 0.07 ± ±<br />
Valine (4.49) ± ± 0.03 ± 0.4 0.7<br />
Total -amino N 3.61 3.74 7.39 3.88 4.4 5.0<br />
Proline 28.25 13.32 28.53 22.05 38.0 40.1<br />
-Amino N <strong>and</strong> proline 31.86 17.06 35.92 25.93 42.4 45.1<br />
Total N 276 336 459 ± ± ±<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.13 Nucleotides in beer (Qureshi et al., 1979)<br />
Concentration ( g/ml) Taste threshold (ppm)<br />
5 0 -Cytidine monophosphate 1.8±7.8 ±<br />
2 0 -Cytidine monophosphate 2.3±7.3 ±<br />
3 0 -Cytidine monophosphate 16.8±73.3 ±<br />
5 0 -Adenosine monophosphate 12.4±67.6 ±<br />
5 0 -Guanosine monophosphate 1.1±4.9 35<br />
3 0 -Adenosine monophosphate 1.6±10.7 ±<br />
5 0 -Uridine monophosphate 2.8±9.1 1.7±4.5<br />
2 0 -Adenosine monophosphate 1.3±5.5 ±<br />
3 0 -Uridine monophosphate 1.5±10.3 ±<br />
2 0 -Uridine monophosphate 9.3±26.2 ±<br />
3 0 -Guanosine monophosphate 1.6±5.1 ±<br />
5 0 -Thymidine monophosphate 2.0±11.9 ±<br />
5 0 -Inosine monophosphate 1.5±4.7 120<br />
3 0 -Inosine monophosphate 1.3±8.1 ±<br />
3 0 -Thymidine monophosphate 1.4±6.6 ±<br />
2 0 -Guanosine monophosphate 2.4±6.7<br />
also examined the small peptides (Mr
Table 9.14 Purines, pyrimidines <strong>and</strong> nucleosides in beer<br />
(Qureshi et al., 1979)<br />
Cytosine (4.58)<br />
Concentration ( g/ml)<br />
11±24.6<br />
Cytidine 18±41<br />
Guanine(4.57) 0.2±3.2<br />
Adenine (4.56) 0.8±5<br />
Uracil (4.59) 1.0±4.6<br />
Uridine 21±70.3<br />
Adenosine 12.5±24.3<br />
Xanthine (4.63) 2.8±9.7<br />
Inosine 1.0±2.4<br />
Guanosine 45±139<br />
Thymidine 7±19.8<br />
Table 19.15 Amides in dark German beer (Tressl et al., 1977)<br />
N,N-Dimethylformamide<br />
N,N-Dimethylacetamide<br />
N-Methylacetamide<br />
N-Ethylacetamide<br />
N-(2-Methylbutyl)acetamide<br />
N-(3-Methylbutyl)acetamide<br />
N-Furfurylacetamide<br />
H.CONMe2<br />
CH3.CONMe2<br />
CH3.CONHMe<br />
CH3.CONH.CH2Me CH3.CONH.CHMe.CH2 CH2Me<br />
CH3.CONH.CH2 CH2. CHMe2<br />
Concentration (ppb)<br />
15<br />
10<br />
+<br />
20<br />
10<br />
25<br />
120<br />
N-(2-Phenylethyl)acetamide CH3.CONH.CH2CH2C6H5 15<br />
(10ppb), have been associated with stale grainy flavours in beer at 2,000 <strong>and</strong> 5ppb<br />
respectively (Palam<strong>and</strong> <strong>and</strong> Grigsby, 1974). Some amides present in beer are given in<br />
Table 19.15 but both they <strong>and</strong> the heterocyclic compounds formed during wort boiling<br />
(Chapter 9) are slightly volatile. Nevertheless, many survive into beer (Table 19.16)<br />
although the concentrations found are usually below the threshold values. Low levels of<br />
pyrroles have also been found in beer (Table 19.17).<br />
Volatile<br />
The biogenic amines <strong>and</strong> polyamines in beer have been reviewed by KalacÏ<strong>and</strong> KrõÂzÏek<br />
(2003). At the pH of beer ammonia <strong>and</strong> the volatile amines found in beer (Table 19.18)<br />
are present astheir non-volatile salts. Ammonia isthe most abundantvolatile nitrogenous<br />
constituent. The mean value in arange of US beers was 14.6ppm (3±33ppm) compared<br />
with 21.3ppm (0±33ppm) in imported beers (Owades <strong>and</strong> Jacevac, 1959). At the low<br />
concentrations found the volatile amines will have little influence on the flavour of beer.<br />
Afterammonia,dimethylamineisthemajorcomponentofthisclass.Tyramine(4.68),the<br />
decarboxylation product of tyrosine, has been detected in beer. Patients being treated for<br />
depression with monoamine oxidase inhibitors must avoid alcoholic drinks <strong>and</strong> yeast<br />
products due to the build up of toxic levels of tyramine. Traces of ethyl carbamate<br />
(urethane, H2N.CO2C2H5), which is reported to be carcinogenic, have been found in<br />
many fermented <strong>and</strong> distilled beverages. In asurvey of 933 samples of Scotch whisky<br />
15±115ppb (mean 43ppb) of ethyl carbamate were found but in asimilar survey of 69<br />
beers, Canas et al. (1989) found that 30 beers had between one <strong>and</strong> four ppb, two had<br />
between 9<strong>and</strong> 13ppb but in the majority ethyl carbamate could not be detected.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.16 Some heterocyclic bases present in beer (Harding et al., 1977)<br />
As mentioned above, secondary amines, such as dimethylamine, can react with oxides<br />
of nitrogen to form carcinogenic N-nitrosamines:<br />
H3C H3C<br />
\ \<br />
2 NH + N2O3 ! 2 N.NO + H2O<br />
/ /<br />
H3C H3C<br />
(4.78) (4.85)<br />
Concentration (ppm)* Threshold in light ale (ppb)<br />
Pyridine (9.33)<br />
2-Acetylpyridine (9.42) 100<br />
3-Acetylpyridine<br />
Pyrazine (9.34) 19±48.5<br />
Methylpyrazine 201±279 (410±419) 1000<br />
2,3-Dimethylpyrazine 0.9±6.5 20<br />
2,5-Dimethylpyrazine 3.4±12.3 50<br />
2,6-Dimethylpyrazine 2.3±16.2 100<br />
Ethylpyrazine 5.9±11.7<br />
2-Ethyl-5-methylpyrazine 5.4±35<br />
2-Ethyl-6-methylpyrazine 0.5±3.4<br />
Trimethylpyrazine 0.7±7.7 100<br />
2-Ethyl-3,5-dimethylpyrazine 6.5±24.6<br />
2-Ethyl-3,6-dimethylpyrazine 1.9±13.8 50<br />
2-Ethyl-5,6-dimethylpyrazine<br />
Tetramethylpyrazine 9.2±62.0 200<br />
Acetylpyrazine 8.1±19.2 100<br />
6,7-Dihydro-5H-cyclopentapyrazine<br />
2-Methyl-6,7-dihydro-5H-cyclopentapyrazine<br />
Thiazole (9.32)<br />
2-Furanmethanol (9.20)<br />
Furfural (9.17)<br />
2-Acetylfuran (9.25) 4±97 80 000<br />
Dihydro-2(3H)furanone<br />
Dihydro-5-methyl-2(3H)furanone<br />
5-Methyl-2-furfural (9.22)<br />
2-Thiophenecarboxaldehyde (9.19)<br />
2-Acetylthiophene (9.27)<br />
5-Methyl-2-thiophenecarboxaldehyde (9.24)<br />
* Qureshi et al. (1979)<br />
N-Nitrosodimethylamine has been found in beer <strong>and</strong> many other foodstuffs. German<br />
beers were found to contain 0±68 ppb (mean 5.9 ppb) N-nitrosodimethylamine while<br />
American beers had 0±14 ppb (mean 5.9 ppb). The highest levels were found in a dark<br />
strong German lager (maximum 47 ppb) <strong>and</strong> in Rauchbier (maximum 68 ppb). N-<br />
Nitrosodiethylamine <strong>and</strong> N-nitrosoproline have also been found in beer. After much<br />
research it was found that the nitrosamines were formed during the kilning of malt<br />
especially in direct fired kilns. Here the mixture of nitrogen oxides, NO x, react with the<br />
amines present, probably hordenine, to form the nitrosamine. With this knowledge the<br />
level of nitrosamines in beer has been greatly reduced. Nevertheless in 1980 the US Food<br />
<strong>and</strong> Drug Administration set a limit of not more than 5 ppb of N-nitrosodimethylamine in<br />
beer <strong>and</strong> the ASBC describe a method to measure nitrosamines in beer.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.17 Pyrrole derivatives in beer (Tressl et al., 1977)<br />
Concentration (ppb)<br />
Pyrrole +<br />
2-Methylpyrrole 1800<br />
2-Formylpyrrole (9.18) 30<br />
2-Acetylpyrrole (9.26) 1400<br />
2-Acetyl-5-methylpyrrole 10<br />
2-Formyl-5-methylpyrrole (9.23) 110<br />
2-Pyrrolidone 10<br />
1-Methyl-2-pyrrolidone +<br />
1-Acetylpyrrole +<br />
1-Furfurylpyrrole 10<br />
Indole +<br />
Table 19.18 Volatile amines present in beer (Slaughter <strong>and</strong> Uvgard, 1971)<br />
Concentration (ppm)<br />
Methylamine (4.64)<br />
Ethylamine(4.65) a<br />
0.03±2.12<br />
n-Propylamine<br />
n-Butylamine<br />
Isobutylamine (4.66) b<br />
0.05±0.10<br />
sec-Butylamine<br />
n-Amylamine<br />
Isoamylamine<br />
Hexylamine<br />
1,3-Diaminopropane<br />
1,4-Diaminobutane (4.74, putrescine)<br />
1,5-Diaminopentane(4.75, cadaverine)<br />
N,N-Dimethyl-1,4-diaminobutane<br />
Dimethylamine (4.78) b<br />
0.07±0.78<br />
Diethylamine<br />
Di-isobutylamine<br />
Pyrrolidine (4.71)<br />
Trimethylamine (4.79) b<br />
0.02±0.06<br />
Tripropylamine<br />
N,N-Dimethylbutylamine<br />
Ethyl carbamate (4.87, urethane) 0±0.013<br />
Ethanolamine<br />
p-Hydroxybenzylamine c (4.80) 0.16±0.72<br />
Tyramine (4.68) 0.15<br />
Histamine (4.69) d<br />
0.08±0.55<br />
a Palam<strong>and</strong> et al. (1969)<br />
b Koike et al. (1972)<br />
c Slaughter <strong>and</strong> Uvgard (1972)<br />
d Chen <strong>and</strong> Van Gheluwe (1979)<br />
19.1.6 Sulphur-containing constituents<br />
Beers contain 100±400ppm of sulphate (Table 19.3). The major non-volatile organic<br />
sulphur compounds in beer are the amino acids cyst(e)ine <strong>and</strong> methionine <strong>and</strong> the<br />
peptides <strong>and</strong> proteins which contain them. Some of these compounds will survive into<br />
beer. In addition malt contains S-methylmethionine (4.157) <strong>and</strong> dimethyl sulphoxide<br />
(4.160) which are precursors of dimethyl sulphide (4.158). Hops may be asource of<br />
sulphur; they may be dusted with elemental sulphur before burr but the burning of<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
sulphur on the oast is less common today. The sulphur compounds in hop oil are<br />
discussed in Chapter 8.<br />
The analysis of volatile sulphur compounds is quite difficult as additional compounds<br />
may be formed if the sample is heated or exposed to light <strong>and</strong>/or oxygen. Headspace<br />
analysis using aflame photometric detector is probably the most satisfactory technique.<br />
Peppard (1985) describes purging beer (sulphur) volatiles, their absorption in aPoropak<br />
Qtrap, <strong>and</strong> their subsequent desorption onto afused silica WCOT capillary column. The<br />
volatile sulphur compounds which have been identified in beer are listed in Table 19.19.<br />
Brewery fermentations can produce up to 10mg/l of sulphur dioxide <strong>and</strong> sodium (or<br />
Table 19.19 Volatile sulphur compounds in beer<br />
Compound Typical levels Flavour Flavour<br />
( g/l) threshold description<br />
( g/l)<br />
Hydrogen sulphide 1±20 5 Sulphidic, rotten eggs<br />
Sulphur dioxide 200±20,000 >25,000 Sulphitic, burnt match<br />
Carbon oxysulphide ± ± ±<br />
Carbon disulphide 0.01±0.3 >50 ±<br />
Thioformaldehyde ± ± ±<br />
Methanedithiol (Dithioformaldehyde) ± ± ±<br />
Thioacetone ± ± ±<br />
Methanethiol 0.2±15 2.0 Putrefaction, drains<br />
Ethylene sulphide 0.3±2.0 20 ±<br />
Ethanethiol 0±20 1.7 Putrefaction<br />
Propanethiol 0.1±0.2 0.15 Putrefaction<br />
1,1-Dimethylethanethiol ± ± ±<br />
2-Furfurylmercaptan ± ± Rubbery<br />
Dimethyl sulphide 10±100 30 Sweetcorn, tin<br />
Diethyl sulphide 0.1±1.0 1.2 Cooked vegetables<br />
Dimethyl disulphide 0.1±3 7.5 Rotten vegetables<br />
Diethyl disulphide 0±0.01 0.4 Garlic, burnt rubber<br />
Dimethyl trisulphide 0.01±0.8 0.1 Rotten vegetables,<br />
onion<br />
n-Butyl methyl sulphide 1 ± ±<br />
Methionol 50±1,300 2,000 Raw potatoes<br />
Methional 20±50 250 Mash potatoes, souplike<br />
Methyl thioacetate 5-20 50 Cabbage<br />
Ethyl thioacetate 0-2 10 Cabbage<br />
3-Methylthiopropionic acid ± ± ±<br />
Ethyl<br />
3-Methylthiopropionate 5±180 ± ±<br />
3-Methylthiopropyl acetate ± ± ±<br />
2-Methyltetrahydro-thiophen-3-one ± ± ±<br />
S-Methyl<br />
2-Methylbutanethiolate ± 1 ±<br />
S-Methyl<br />
4-Methylpentanethiolate ± 15 ±<br />
S-Methyl hexanethiolate ± 1 ±<br />
4-(4-Methylpent-3-enyl)-3,6-dihydro-<br />
1,2-dithiine ± ± ±<br />
3-Methyl-2-butene-1-thiol 0.001±0.1 0.01 Skunk, leek-like, light<br />
struck<br />
After Baxter <strong>and</strong> Hughes (2001) with additions<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
potassium) metabisulphate may be added as apreservative either during conditioning or<br />
at racking. However, EEC regulations (1987) have limited the level of SO2 in packaged<br />
beer to 20mg/kg <strong>and</strong> to 50mg/kg in cask conditioned beers. In contrast, 160±400mg/kg<br />
SO 2areallowedinwine<strong>and</strong>200mg/kgincider.Muchofthesulphurdioxideinbeerisin<br />
abound form such as the acetaldehyde bisulphite compound. There are several approved<br />
methods for the analysis of sulphur dioxide in beer. Analytica-EBC <strong>and</strong> the Institute of<br />
<strong>Brewing</strong> use the classicMonier-Williams distillation method inwhich the SO2isdistilled<br />
in acurrent of nitrogen or carbon dioxide into hydrogen peroxide <strong>and</strong> the sulphuric acid<br />
formed titrated with st<strong>and</strong>ard sodium hydroxide. The ASBC describe amethod whereby<br />
the colour which is restored to acid-decolourised p-rosaniline hydrochloride is measured.<br />
In a collaborative trial (Baker <strong>and</strong> Upperton, 1992) these two methods gave better<br />
precision than the IoB rapid method or the EBC dithiobisnitrobenzoic acid method.<br />
However, concern has been expressed about the possible carcinogenic nature of prosaniline.<br />
Analytica-EBC also describes an enzymatic method for sulphur dioxide. Sulphite is<br />
oxidized to sulphate by the enzyme sulphite oxidase (SO2.OD):<br />
SO2.OD<br />
SO3 2 ‡O2 ‡H2O !SO4 2 ‡H2O2<br />
The hydrogen peroxide produced is reduced by the enzyme NADH-peroxidase (NADH-<br />
POD):<br />
NADH-POD<br />
H2O2 ‡NADH‡H ‡<br />
!NAD ‡ ‡2H2O<br />
<strong>and</strong> the NADH is measured by its absorption at 340nm. The level of SO2 in beer seldom<br />
exceeds the flavour threshold <strong>and</strong> declines during storage with ahalf life between 37 <strong>and</strong><br />
221 days; in one beer ahalf life of 1,038 days was found (Ilett et al., 1996).<br />
The hydrogen sulphide present inbeer is also partlyin abound form but the total level<br />
may exceed the threshold level. At this low level the flavour impact is not unpleasant <strong>and</strong><br />
is characteristic of the flavour of ales especially cask conditioned ales which may have<br />
had potassium metabisulphite added as apreservative. Lager yeasts are less efficient in<br />
reducing SO2 to H2S than ale yeasts. In avigorous fermentation much of the H2S will be<br />
removed with the carbon dioxide. Yeast autolysis or microbial infection, e.g., by<br />
Zymomonas spp., can also produce hydrogen sulphide. The amounts of methanethiol,<br />
ethanethiol <strong>and</strong> propanethiol in beer (Table 19.19) probably exceed their flavour<br />
thresholds.<br />
Dimethyl sulphide (4.158, DMS) is an important component of the flavour of lager<br />
beers. The concentration in arange of beers, given in Table 19.20, shows the higher<br />
levels in lager beers. As mentioned above it is mainly formed by the breakdown of Smethylmethionine<br />
present in malt. Lightly kilned lager malts will retain more Smethylmethionine<br />
than ale malts. Some brewers specify the level of S-methylmethionine<br />
in the malt they buy. DMS formed during kilning <strong>and</strong> wort boiling will be lost to the<br />
atmosphere but that formed in the whirlpool <strong>and</strong> during later processing is likely to<br />
remain in the wort unless vapour stripping is employed. Some DMS will be purged with<br />
the CO2 during fermentation but some will survive into beer. During kilning some Smethylmethionine<br />
is converted into dimethyl sulphoxide (DMSO) which is not as volatile<br />
as DMS but soluble in water so it will survive into wort. Here some yeast strains are<br />
capable of reducing it back to dimethyl sulphide. Late or dry hopping with whole hops,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.20 Dimethlyl sulphide content of beer (ppb)<br />
British ales 14 Sinclair et al. (1970)<br />
British lagers 16±27 Sinclair et al. (1970)<br />
Continental lagers 44±114 Sinclair et al. (1970)<br />
Beer from green malt 80 Anderson et al. (1975)<br />
German lagers 32±205 (av. 94) Narziss et al. (1979)<br />
Diet beer 46±98 (av. 63.5) Narziss et al. (1979)<br />
Canadian ale 92 Hysert et al.(1979)<br />
Canadian lager 114 Hysert et al. (1979)<br />
Canadian low-alcohol beer 82 Hysert et al. (1979)<br />
British 41±75 ±<br />
German 141±153 ±<br />
United States 59±106 ±<br />
will add as much as 15 ppb to the beer dimethyl sulphide level (Moir, 1994). Hegarty et<br />
al. (1995) found that 2-phenylethanol can suppress the flavour intensity of dimethyl<br />
sulphide. Dimethyl sulphide in beer is usually estimated by headspace gas chromatography<br />
with acapillary column. Aflame photometric detector is specific for sulphur<br />
compounds but aflame ionization detector gives similar results for DMS <strong>and</strong> also allows<br />
measurement of acetaldehyde, ethyl acetate, n-propanol, isobutanol <strong>and</strong> isoamyl alcohol<br />
(Analytica-EBC, IoB, Dupire, 1999) (see also Mundy, 1991). 3-Methylthiopropionaldehyde<br />
was found to contribute the worty flavour to alcohol-free beers (Perpete <strong>and</strong> Collin,<br />
1999). The same compound is the precursor of dimethyl trisulphide.<br />
As discussed in Chapter 8, photolysis of the iso- -acids in the presence of asensitizer<br />
(riboflavin) <strong>and</strong> asuitable sulphur donor can produce alight- or sun-struck flavour. This<br />
is due to 3-methyl-2-butene-1-thiol (8.48, prenyl mercaptan) which produces askunklike,<br />
leek-like aroma with an extremely low threshold of 10ng/l (0.10ppb). To avoid this<br />
off-flavour few beers are now bottled in clear glass bottles. Alternatively the beer may be<br />
bitteredwitheithertetrahydroiso- -acidsor -iso- -acidswhicharenotaffectedbylight.<br />
19.2 Nutritive value of beer<br />
The nutritional aspects of beer are discussed by Baxter <strong>and</strong> Hughes (2001). The calorific<br />
(caloric, energy) value of beer is calculated from the alcohol, carbohydrate <strong>and</strong> protein<br />
contents. The (UK) Food Labelling Regulations (1980) give the following figures for<br />
calculating calorific values: 1gcarbohydrate (as monosaccharide) ˆ3.75 kilocalories,<br />
1gprotein ˆ4kcal, 1gfat ˆ9kcal, <strong>and</strong> 1galcohol ˆ7kcal. Accordingly the IoB give<br />
the formula:<br />
Calorific value …kcal=100ml† ˆ7…A†‡3:75…C†‡4…P†<br />
where Aˆalcohol content/100 ml, Cˆtotal carbohydrate (as glucose)/100ml, <strong>and</strong> Pˆ<br />
protein content/100ml (Martin, 1982). Alternatively, the result may be expressed in<br />
kilojoules (1kcal ˆ4.184kJ):<br />
Energy value …kJ=100ml† ˆ29…A†‡17…C†‡17…P†<br />
Some typical energy values are given in Table 19.21. It can be seen that alcohol<br />
contributes more energy than the carbohydrates so low carbohydrate `lite' beers can have<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.21 Calorific (energy) values of beer<br />
Beer type Energy Value<br />
kJ/100 ml kcals/100 ml<br />
`Lite' beers 75±110 20±26<br />
Lagers 85±125 20±30<br />
Ales 114±160 25±38<br />
Stouts/strong beers 150±300 35±70<br />
From Baxter <strong>and</strong> Hughes (2001)<br />
Table 19.22 Vitamin content of beers<br />
Vitamin<br />
(mg/l)<br />
Range in beers (mg/l) Typical level in UK beer<br />
Niacin 3±20 7.7<br />
Riboflavin 0.07±1.3 0.3<br />
Pyridoxine (B6) 0.13±1.7 0.5<br />
Foliates 0.03±0.10 0.05<br />
Biotin 0.007±0.018 0.01<br />
B12 0.09±0.14 0.1<br />
Pantothenic acid 0.5±2.7 0.9<br />
Thiamine<br />
After Baxter <strong>and</strong> Hughes (2001)<br />
0.002±0.14 0.03<br />
high energy values due to the alcohol present. However, in some countries the term `lite'<br />
refers to low-alcohol beers.<br />
The ASBC use slightly different figures to calculate the caloric content of beers (i.e. 1<br />
gcarbohydrate ˆ4kcal <strong>and</strong> 1g alcohol ˆ6.9kcal) <strong>and</strong> take into account the ash content<br />
of the beer so that:<br />
Caloric content …kcal=100ml beer† ˆ6:9…A†‡4…B C†<br />
where AˆAlcohol (% by weight), BˆReal extract (% by weight), <strong>and</strong> Cˆash content<br />
(%byweight).Beerisarichsourceofcertain vitamins(Table19.22).Baxter<strong>and</strong>Hughes<br />
(2001) discuss the contribution that beer can make to the required daily intake of the<br />
vitamins.<br />
19.3 Colour of beer<br />
The physical properties assessed by abeer drinker include colour, clarity, viscosity <strong>and</strong><br />
foam. Obviously if the beer is drunk directly from acan or bottle these properties will<br />
have less impact on the consumer. The colour of beer is largely due to the melanoidins<br />
<strong>and</strong> caramel present in the malt <strong>and</strong> adjuncts used but further caramelization can take<br />
place during wort boiling (see Chapters 9<strong>and</strong> 10). Minor adjustments of the colour of<br />
beer can be made by the addition of caramel either to the copper or with the primings.<br />
Other contributors to the colour of beer are oxidized polyphenols especially in the<br />
presence of trace metals such as iron or copper. In pale beers the yellow vitamin<br />
riboflavin (
Extinction<br />
Wave numbers (×10<br />
25 24 23 22 21 20 19 18 17 16 15<br />
–3 ) cm –1<br />
Wave numbers (×10<br />
25 24 23 22 21 20 19 18 17 16 15<br />
–3 ) cm –1<br />
(a) (b)<br />
0.9<br />
0.8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
f<br />
b<br />
c<br />
d<br />
a<br />
A new series of glass discs were released in 1950 <strong>and</strong> adopted by the EBC. The discs (2±<br />
27 EBC colour units) were matched against the beer in either 5, 10, 25 or 40 mm cells.<br />
However, it is necessary that every person entrusted to the measurement of the colour of<br />
worts <strong>and</strong> beers in this way is known to be free from colour blindness: 10% of the male<br />
<strong>and</strong> 1% of the female population do not have perfect colour vision. Ishihara (1964) has<br />
provided tests for colour blindness. In view of these difficulties spectrophotometric<br />
measurements at one wavelength were adopted. Analytica-EBC chose measurements at<br />
430 nm in a 1 cm cell when:<br />
EBC Colour ˆ A430 25<br />
Originally the Institute of <strong>Brewing</strong> chose measurements at 530 nm as being more<br />
suited to ales but later adopted 430 nm. The ASBC also take measurements at 430 nm but<br />
in a half-inch cell when:<br />
Color …ASBC† ˆ A430 10<br />
e<br />
Key<br />
a Pale ale 1<br />
b Pale ale 2<br />
c Pale ale 3<br />
d Light ale 1<br />
e Light ale 2<br />
f Pale ale 4<br />
417 435 455 477 500 526 556 589 625 667<br />
Wavelength (nm)<br />
Nevertheless, single wavelength measurements can provide only limited information. In<br />
the human eye cone cells in the centre of the retina respond to either red (c. 600 nm),<br />
green (c. 550 nm) or blue (c. 450 nm) light <strong>and</strong> send three signals to the brain which are<br />
there integrated <strong>and</strong> interpreted as colour. Hence tristimulus values are being increasingly<br />
used to measure the colour of beers (Sharpe et al. 1992, Smedley, 1992, 1995).<br />
When evaluating colour the energy of the illuminant light source across the visible<br />
spectrum (`colour temperature') must be taken into consideration. The Commission<br />
Internationale de l'Eclairage (CIE) have defined a number of st<strong>and</strong>ard illuminants:<br />
Illuminant B (colour temperature, 4,900 K) represents sunlight, Illuminant C (6,800 K)<br />
represents average daylight, <strong>and</strong> Illuminant D65, which is most commonly used,<br />
represents daylight with some UV correction. The CIE defines colour in terms of three<br />
parameters: hue (hë), value (L*), <strong>and</strong> chroma (C*) as illustrated in the CIELAB colour<br />
Extinction<br />
0.8<br />
0.7<br />
0.6<br />
0.5<br />
0.4<br />
0.3<br />
0.2<br />
0.1<br />
0<br />
e<br />
d<br />
c<br />
b<br />
a<br />
Key<br />
a-d Largers<br />
e Stout<br />
417 435 455 477 500 526 556 589 625 667<br />
Wavelength (nm)<br />
Fig. 19.2 (a) Spectra of commercial ales; (b) spectra of lager <strong>and</strong> stout (Hudson, 1969).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
space (Fig. 19.3). `Hue' is the term ascribed to what we generally consider as the<br />
prominent `colour' of an object. The hues, red, yellow, green, blue <strong>and</strong> purple form a<br />
continum in the horizontal plane of the colourspace which is sometimes called the colour<br />
wheel (Sharpe et al., 1992). It should be noted that the hue (h) is a measure of angular<br />
rotation. The value (L*) is a measurement of `light' <strong>and</strong> `dark' <strong>and</strong> is a vertical<br />
displacement from the north pole (white) to the south pole (black). The chroma (C*)<br />
measures what observers call `dull' or `vivid' <strong>and</strong> is a horizontal displacement of the<br />
vertical axis of the colour space towards the circumference. For dealing with differences<br />
in colour between two samples it is generally more convenient to express locations within<br />
the colour space using Cartesian co-ordinates. Here, hue <strong>and</strong> chroma are combined into<br />
two parameters a* <strong>and</strong> b* so:<br />
<strong>and</strong><br />
Chroma …C † ˆ a 2<br />
‡ b 2<br />
Hue …h † ˆ tan 1 …b =a †<br />
Commercial tristimulus transmission instruments are available which record colour in<br />
terms of L*, a* <strong>and</strong> b*. There is a linear relationship between L* <strong>and</strong> EBC colour units<br />
but tristimulus measurements can distinguish between beers showing the same EBC<br />
colour values (Smedley, 1995). The ASBC have recommended the use of tristimulus<br />
analysis for measurement of beer colour (Cornell, 2002).<br />
19.4 Haze<br />
Blue<br />
Green<br />
Chroma<br />
White<br />
Black<br />
Value<br />
Yellow<br />
Fig. 19.3 A representation of the CIELAB colour space showing the relationship between<br />
value, hue <strong>and</strong> chroma (Smedley, 1995)<br />
Clarity is the absence of haze. Most drinkers expect their beer to be bright <strong>and</strong> clear <strong>and</strong><br />
may reject cloudy beer untasted. Beer hazes are of two types: biological <strong>and</strong> nonbiological.<br />
Infection of bright beer with either bacteria or wild yeasts will produce a<br />
biological haze due to the growth of the invading organism when the beer will usually<br />
become sour <strong>and</strong> unacceptable. With the use of pasteurization <strong>and</strong> sterile filtration,<br />
Hue<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Red
iological infection <strong>and</strong> haze formation is fairly rare. However, sterile beers kept for a<br />
length of time will develop anon-biological haze. The rate of development of such hazes<br />
determinestheshelf-lifeofbottled<strong>and</strong>cannedbeers.Beforebeershowsanyhazeatroom<br />
temperature (20ëC) it may form achill haze if cooled to 0ëC, Such chill hazes will<br />
redissolve at 20ëC when permanent hazes will remain. Chill hazes are obviously more of<br />
aproblem with lager beers, which may be served as cold as 4ëC, than with ales. Beer<br />
haze has been reviewed by Bamforth (1999) <strong>and</strong> by Siebert (1999).<br />
19.4.1 Measurement of haze<br />
The amount of chill haze isolated for an EBC collaborative analysis from three different<br />
breweries was between 1.4 <strong>and</strong> 8.1mg/l while the permanent haze accounted for 6.6±<br />
14.6mg/l (Carrington et al., 1972). The amount of permanent haze increases with time of<br />
storage <strong>and</strong> yields of up to 44mg/l have been reported but beers are commercially<br />
unacceptable long before the haze can be measured gravimetrically. When light is passed<br />
through asuspension of acoloured precipitate <strong>and</strong> the amount of reflection is negligible,<br />
the light absorption gives ameasure of the turbidity according to Lambert's law. With<br />
white precipitates, such as beer haze, much of the light is reflected so measurements are<br />
made of the light reflected at agiven angle to the incident light (nephelometry). The<br />
angle chosen <strong>and</strong> the size of the haze particles are the two most important factors which<br />
determine the amount of haze perceived (Thorne <strong>and</strong> Nannested, 1960). Most of the<br />
commercial instruments in use today take measurements at 90ëto the incident light: the<br />
exception is the Monitek 251 which takes readings at aforward angle of 13ë. The<br />
wavelength of the incident light also varies between instruments (350±860nm) <strong>and</strong> some<br />
instruments are more sensitive to colour than others (Buckee et al., 1986). Mundy <strong>and</strong><br />
Boley (1999) concluded that there are significant differences between the haze values<br />
obtained on the same sample with different instruments, confirming that the results<br />
obtained on different instruments cannot be directly compared. The EBC, IoB <strong>and</strong> ASBC<br />
all use formazin, prepared by the reaction between hydrazine sulphate <strong>and</strong><br />
hexamethylenetetramine, as the primary haze st<strong>and</strong>ard. Unfortunately the EBC <strong>and</strong> the<br />
ASBC have adopted different scales so:<br />
10;000 ASBC Formazin Turbidity units ˆ145 EBC Formazin haze units<br />
1EBC haze unit ˆ69 ASBC haze units<br />
Earlier, barium sulphate (Helm) <strong>and</strong> fullers' earth were used as st<strong>and</strong>ards. A<br />
comparison of the different scales with visual assessment is given in Fig. 19.4. Styrenedivinylbenzene<br />
copolymer suspensions (AEPA) have also been used as turbidity<br />
st<strong>and</strong>ards but samples from different suppliers had different mean particle sizes. Coulter<br />
counter analyses showed that the majority of the particles in the Formazin haze st<strong>and</strong>ard<br />
were between 1.5 m <strong>and</strong> 2.5 m diameter (mean 2.1 m) (Morris, 1987). In most cases<br />
visual assessment of beer haze correlates well with instrument readings for light scattered<br />
at 90ë but some beers which appear bright to the eye give substantial meter readings.<br />
Such beers were said to contain `invisible' or `pseudo' haze. These `pseudo' hazes were<br />
not observed with instruments using 13ë forward light scattering. Morris (1987)<br />
investigated the relationship between haze <strong>and</strong> particle size. He found that the 90ë<br />
(Radiometer) haze meter gave results which can be closely related to turbidity for<br />
particles > 0.5 m in diameter but for particles < 0.5 m it was somewhat oversensitive.<br />
A high reading will therefore be obtained for a suspension containing particles < 0.5 m<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Comparison of haze scales<br />
It must be stressed again that no exact equivalence must be expected between different<br />
haze scales or between visual observations made in different laboratories. However, a rough<br />
comparison between haze scales can be made with the help of this nomogram <strong>and</strong> with the<br />
relationship:<br />
1 EBC Formazin Haze Unit = 40 Helm units = 69 ASBC Formazin units<br />
10<br />
EBC<br />
units<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
Helm<br />
units<br />
400<br />
350<br />
300<br />
250<br />
200<br />
150<br />
100<br />
50<br />
ASBC<br />
units<br />
700<br />
Verbal<br />
assessment<br />
diameter if the calibration was originally carried out with particles > 0.5 m, which<br />
explains the observation of `invisible' or `pseudo' haze.<br />
19.4.2 Composition <strong>and</strong> formation of haze<br />
The most common non-biological hazes found in beer are formed by interactions between<br />
proteins <strong>and</strong> polyphenols. Chill hazes contain between 45.5 <strong>and</strong> 66.8% of proteins, the<br />
hydrolysates of which are rich in glutamic acid, proline, arginine <strong>and</strong> aspartic acid<br />
residues. Alkaline hydrolysis of beer hazes produces a range of phenolic acids including:<br />
ferulic, sinapic, vanillic, syringic, gallic, protocatechuic <strong>and</strong> caffeic acids (Harris, 1965).<br />
In contrast, acid treatment of beer hazes liberates the pigments cyanidin <strong>and</strong> delphinidin<br />
showing the presence of proanthocyanidins in beer haze. However, none of the<br />
proanthocyanidins found in malt, hops <strong>and</strong> beer contain a methoxy group so the ferulic,<br />
vanillic, sinapic, <strong>and</strong> syringic acids produced on alkaline hydrolysis must come from<br />
another source probably lignin. Barley straw lignin contains 16.4% methoxyl so, on the<br />
assumption that all the methoxyl in haze is derived from lignin <strong>and</strong> that this lignin has the<br />
same composition as barley straw lignin, hazes contain 5.7±7.9% lignin (Harris, 1965).<br />
Hazes contain 0.7±3.3% of ash rich in many metals (Hudson, 1955). Some of the metals,<br />
e.g., copper, iron <strong>and</strong> aluminium, are concentrated in haze up to 80,000 times the level in<br />
the residual beer. Hazes also contain 2±4% of glucose <strong>and</strong> traces of the pentoses<br />
arabinose <strong>and</strong> xylose.<br />
600<br />
500<br />
400<br />
300<br />
200<br />
100<br />
Very hazy<br />
Hazy<br />
Slightly hazy<br />
Very slighly hazy<br />
A1 brilliant<br />
Brilliant<br />
Fig. 19.4 Comparison of haze scales (Analytica-EBC, 1998).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
During wort boiling (Chapter 9) some protein is removed in the hot break or trub but<br />
protein-polyphenol complexes, which dissociate at 80ëC, are not found in the hot break but<br />
do occur in the cold break. Proteins not removed during wort boiling may survive<br />
fermentation <strong>and</strong> persist into beer where they may cause haze. However, not all proteins<br />
nor all polyphenols are haze-active. The suggestion that hydrophilic proteins are<br />
responsible for haze <strong>and</strong> hydrophobic proteins are necessary for head retention is probably<br />
too facile. Haze-active proteins or polypeptides in beer are mainly derived from the barley<br />
prolamins or hordeins, which are alcohol soluble <strong>and</strong> rich in proline. Indeed proline<br />
residues appear to be necessary for haze formation. In amodel system with catechin <strong>and</strong><br />
different polypeptides, haze increased linearly with the mole %proline in the polypeptide,<br />
the most haze being formed with polyproline. Polypeptides which contain little or no<br />
proline produce little or no haze. However, Ishibashi et al. (1996) found that antibodies<br />
raised against foaming polypeptides were specific for those compounds but antibodies<br />
raised against haze reacted with both haze <strong>and</strong> foaming polypeptides. It is suggested that<br />
albumin- <strong>and</strong> globulin-derived material may adhere to particles formed originally from<br />
hordein-derived polypeptides. Yang <strong>and</strong> Siebert (2001) investigated dyes that bind to<br />
proteins <strong>and</strong>foundthatbromopyrogallolredgavethebestindicationofhaze-activeprotein.<br />
Similarly, not all the polyphenols present in beer are haze-active; proanthocyanidins<br />
<strong>and</strong> flavanols are the most important. Catechin, epicatechin <strong>and</strong> gallocatechin give small<br />
amounts of haze with abeer haze-active protein, but the dimers, e.g., procyanidin B-3,<br />
produce much more haze <strong>and</strong> the trimers more again. As mentioned above, tetramers <strong>and</strong><br />
pentamers do not survive into beer but they may reform by oxidation as beers age. Of the<br />
dimers, prodelphinidin B-3 produces more haze than procyanidin B-3. The amount of<br />
haze formed depends on the concentrations of both the protein <strong>and</strong> the polyphenol <strong>and</strong><br />
their ratio. Siebert et al. (1996) produced amodel for protein-polyphenol interactions<br />
(Fig. 19.5) in which both the polyphenols <strong>and</strong> the haze-active proteins have afixed<br />
number of binding sites (shown as two <strong>and</strong> three respectively in Fig. 19.5). When the<br />
number of polyphenol `ends' equals the number of protein binding sites, the largest<br />
network with the largest particles will be formed which will show the greatest amount of<br />
lightscattering.Withanexcessofprotein,thepolyphenolwillformabridgebetweentwo<br />
peptide chains but there will be insufficient for further bridges. With an excess of<br />
polyphenol relative to protein, all the protein binding sites will be occupied <strong>and</strong> it is<br />
unlikely that the free end of the polyphenol will find avacant protein site for further<br />
cross-linking.<br />
Occasionally hazes which differ from the normal protein-polphenol haze are found in<br />
beer. These include calcium oxalate hazes (the solubility of calcium oxalate is only<br />
6.07mg/l at 13ëC) <strong>and</strong> carbohydrate hazes such as retrograded starch ( -glucan), -<br />
glucans <strong>and</strong> pentosans (Chapter 15).<br />
19.4.3 Prediction of haze <strong>and</strong> beer stability<br />
The conditions to which any particular beer package will be subjected cannot be known<br />
beforeh<strong>and</strong>; therefore two types of test are adopted by most breweries:<br />
1. Long-term storage at a temperature related to that likely in trade with an examination<br />
for biological <strong>and</strong> non-biological haze at the end of a defined period (e.g. three or six<br />
months).<br />
2. Accelerated haze production under defined high temperature conditions, to give an<br />
early indication of the liability to non-biological haze.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
[Polyphenol] = [Protein]<br />
[Polyphenol] > [Protein]<br />
Polyphenol molecule<br />
Protein molecule with fixed<br />
number of polyphenol binding<br />
sites (i.e. haze-active)<br />
Fig. 19.5 Conceptual mechanism of protein-polyphenol interaction (Siebert et al., 1996).<br />
Copyright (1996) American Chemical Society<br />
The EBC recommendation is to measure the increase in non-biological haze at the end of<br />
a defined period at a defined temperature. Brewers will adopt different conditions to suit<br />
their beers but for comparison the EBC recommends that six bottles should be cooled to<br />
0 ëC overnight <strong>and</strong> the initial haze read in the morning at 0 ëC. The bottles are then placed<br />
in an upright position <strong>and</strong> without agitation, in a bath at 60 ëC for 48 hours. At the end of<br />
this time they are again cooled at 0 ëC overnight <strong>and</strong> the haze measured the following<br />
morning at 0 ëC The initial <strong>and</strong> final haze values are reported in EBC Formazin units.<br />
The ASBC measure the total haze after chilling in a similar manner <strong>and</strong> describe<br />
accelerated chill haze tests. They suggest that 12 weeks or three months is a reasonable<br />
time in which beer can be expected to be sold. Accordingly, they recommend storing 24<br />
bottles or cans at 22 ëCÔ2 ëC in a vertical position without agitation for eight weeks.<br />
Three containers are then cooled overnight at 0 ëC <strong>and</strong> the haze measured at 0 ëC. Each<br />
week, up to 12 weeks, three more containers are cooled <strong>and</strong> the haze measured. For the<br />
accelerated chill haze test three containers are held at the selected temperature (40, 50, or<br />
60 ëC) for one week. The samples are then cooled at 0 ëC for 24 hours <strong>and</strong> the haze<br />
measured. (Note that caution should be exercised when the test is run at 60 ëC because<br />
there is some danger that bottles may burst <strong>and</strong> cans may rupture.) The results of the<br />
accelerated chill haze tests, in Formazin Turbidity Units, are compared with the haze<br />
formed after 12 weeks at 22 ëC to determine which forcing temperature gives results that<br />
correlate best with room temperature storage. Thereafter only that forcing temperature<br />
need be used routinely.<br />
These accelerated shelf-life test take at least 48 hours or one week to give results.<br />
More rapid predictions of beer stability are sometimes obtained using chemical<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
precipitants. The IoB describe methods for sensitive protein in beer, alcohol chill haze in<br />
beer, <strong>and</strong> for the saturated ammonium sulphate precipitation limit of beer.<br />
19.4.4 Practical methods for improving beer stability<br />
Brewers wish to increase the shelf-life of their beers <strong>and</strong> adopt various strategies to<br />
achieve this, i.e., removal of protein, removal of polyphenols or removal of aportion of<br />
each (Chapter 15). This last method is usually employed since some proteins <strong>and</strong> some<br />
polyphenols are necessary for the character of the beer. To reduce the level of proteins<br />
some (lager) brewers used to treat their beers with proteolytic enzymes (papain,<br />
bromelain,ficin,orenzymesfromBacillussubtilis)butsuchtreatmentislikelytoremove<br />
both haze-active <strong>and</strong> foam-active proteins. Similarly, treatment with bentonite, removes<br />
both types of protein but silica hydrogels are more specific for haze-active proteins<br />
(Siebert <strong>and</strong> Lynn, 1997). Apperson et al. (2002) studied the silica absorption of hazeactive<br />
beer proteins <strong>and</strong> found that non-activated small pore volume silica (silica B) was<br />
one of the most efficient absorbers of isolated haze protein. Alternatively, tannic acid is a<br />
relatively efficient precipitant of haze-active proteins in beer.<br />
Proanthocyanidin-free barley varieties have been bred <strong>and</strong> malted which, when used<br />
with apolyphenol-free hop extract, give colloidally stable beers. However, most brewers<br />
still rely on absorbants to remove excess polyphenols. After nylon <strong>and</strong> polyvinylpyrrolidone<br />
(PVP), cross-linked polyvinylpyrrolidone, polyvinylpolypyrrolidone (PVPP) is<br />
now the agent of choice (McMurrough, 1995) often in conjunction with silica hydrogels.<br />
Two types of PVPP are available. The first, single-use PVPP which is amicroized white<br />
powder with ahigh surface/weight ratio, is used in afilter aid bodyfeed dosing regimen.<br />
The second type is regenerable PVPP used either in impregnated sheets or within a<br />
horizontal leaf pressure vessel. These materials are regenerated by treatment with 1±2%<br />
sodium hydroxide, followed by washing <strong>and</strong> neutralization. According to Bamforth<br />
(1999), this last treatment is the cheapest. Siebert <strong>and</strong> Lynn (1998) compared polyphenol<br />
interactions with PVPP <strong>and</strong> haze-active protein. They concluded that the mechanisms by<br />
which haze-active polyphenols attach to PVPP <strong>and</strong> to haze-active proteins are similar but<br />
not identical.<br />
19.5 Viscosity<br />
Dynamic viscosity is defined as the resistance to shear flow within aliquid. Kinematic<br />
viscosity is ameasure of the time taken by aliquid to flow through an orifice under<br />
gravity. Both are measured by an international method using the time of flow in an<br />
Ostwald viscometer at 20.0ëC. At this temperature the dynamic viscosity of water is<br />
1.002cP (centipoises) <strong>and</strong> the kinematic viscosity is 1.00cS (centistokes). The SI unit for<br />
dynamic viscosity is the Pascal-second when 1cP ˆ0.001Pa.s. The EBC use a20% w/v<br />
sucrose solution as an additional st<strong>and</strong>ard (1.945 cP) but the Institute of <strong>Brewing</strong>,<br />
following BS188: 1957, use a 3.0% w/v solution of glycerol. Some values are: worts (SG<br />
1,030±1,100) 1.59±5.16, lager (SG 1,007) 1.45; <strong>and</strong> stout (SG 1,009) 1.96 cP. The method<br />
can be extended to measure the viscosity of liquid sugars <strong>and</strong> brewing syrups using a<br />
viscometer of the appropriate range.<br />
The viscosity of wort <strong>and</strong> beer is influenced by the macromolecules present, Sadosky<br />
et al. (2002) found that arabinoxylan, -glucans <strong>and</strong> dextrins all increased the viscosity of<br />
model solutions with the dextrins having the largest effect.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
19.6 Foam characteristics <strong>and</strong> head retention<br />
The vast literature on this subject has been reviewed by Bamforth (1985), Baxter <strong>and</strong><br />
Hughes (2001) <strong>and</strong> Evans <strong>and</strong> Sheehan (2002). The European Brewery Convention<br />
(1999) have published the proceedings of a symposium on Beer Foam Quality, held in<br />
Amsterdam in 1998. Bamforth (1985) defined beer foam quality as a combination of its<br />
stability, quantity, lacing (adhesion or cling), whiteness, `creaminess', <strong>and</strong> strength.<br />
Further, he provided evidence (2000) that consumers differ, along regional <strong>and</strong> gender<br />
lines, in their requirements for the amount of beer foam, its stability <strong>and</strong> whether it laces<br />
the glass. In Engl<strong>and</strong>, draught beers from Burton-on-Trent are served with little or no<br />
foam but in the north-east a rich creamy head that overflows the glass is expected. This<br />
can cause problems when the glass is a legal measure. However, the judgment of UK law<br />
is that the customers' preference should be catered for. Thus the head can be considered<br />
an integral part of the pint or the drinker can dem<strong>and</strong> a full pint of liquid.<br />
Beer foams are colloidal systems comprising a continuous liquid phase <strong>and</strong> a<br />
discontinuous gas phase; the bulk density of the system approaches that of a gas rather<br />
than that of a liquid. Beer foam physics have been discussed by Ronteltap et al. (1991)<br />
<strong>and</strong> by Prins <strong>and</strong> van Marle (1999). Ronteltap et al. (1991) concluded that foam<br />
characteristics are determined by four key processes: bubble formation, drainage,<br />
coalescence, <strong>and</strong> disproportionation. Bubble formation requires a suitable site for<br />
nucleation. Bubbles may be generated in beer either by dispersal or condensation. The<br />
simplest dispersal system involves injection of a gas from a capillary. A spherical bubble<br />
forms at the tip of the capillary <strong>and</strong> will be released when its buoyancy is greater than the<br />
surface tension effects that hold the bubble to the tip of the capillary. Bubbles formed in<br />
this way are of similar size whereas those generated from a sinter will be more<br />
heterogeneous <strong>and</strong> more likely to undergo diproportionation. Homogeneous condensation<br />
occurs, for example, when a crown cork is removed from a bottle of beer at, say, 5 ëC,<br />
when rapid gas expansion may cause the temperature to drop locally as low as 36 ëC.<br />
Much more likely is heterogeneous condensation when the gas already present (usually<br />
air in the case of beer dispense) is exp<strong>and</strong>ed by diffusion of gas (either carbon dioxide or<br />
nitrogen) from the solution into the gas phase. As the bubble grows it experiences greater<br />
buoyancy until it breaks away from the nucleating surface leaving a small pocket of<br />
residual gas to begin the process again. So the nature <strong>and</strong> amount of gas in solution<br />
influence the beer foam (Fisher, et al., 1999) as does the angle of dispense.<br />
Etched glassware <strong>and</strong> small particles may provide nucleation sites but lose their<br />
effectiveness when totally wetted or not scrupulously clean. Agitation of beer can<br />
produce nucleation sites by cavitation, the instantaneous formation of a vacuum when the<br />
liquid separates from the vessel wall. Ultrasound can promote cavitation but often with<br />
uncontrollable gushing. Bubble formation involves an increase in the surface area of the<br />
liquid which is opposed by the surface tension of the liquid but bubbles remain stable<br />
because of materials of low surface tension within the bubble wall which have both<br />
hydrophobic <strong>and</strong> hydrophilic areas (e.g. proteins). Once a bubble leaves its nucleation site<br />
it will rise through the beer until it reaches the beer-air/foam interface. Surface active<br />
materials in the bubble wall may facilitate this movement. The nature of the gas within<br />
the bubble is important <strong>and</strong> the pressure within the bubble is inversely proportional to its<br />
diameter so gas will pass from smaller bubbles to larger ones (disproportionation). If the<br />
gas is soluble in the liquid film the passage of gas between the bubbles is faster. Thus<br />
foams containing oxygen <strong>and</strong> nitrogen produce smaller bubbles <strong>and</strong> a more stable foam<br />
than those containing carbon dioxide, which is more soluble in the liquid film. As the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
spherical bubble leaves the beer-foam interface it will be pushed upwards by younger<br />
bubbles arriving from below. As it moves to the top of the head the liquid between the<br />
individual bubbles starts to drain away, the fresh wet foam is converted to dry foam <strong>and</strong><br />
finally only asolid network of bubble walls remains which may be deposited from the<br />
liquid phase on to the glass showing cling, lacing, or foam adhesion. To the consumer<br />
foam drying or drainage results only in amodest change in foam volume but bubble<br />
coalescence or disproportionation, with the formation of larger coarser bubbles, is amore<br />
obvious sign of foam destabilization.<br />
19.6.1 Methods of assessing foam characteristics<br />
Methods have been devised for measuring many foam parameters such as foamability,<br />
foam stability, foam drainage, cling, viscoelasticity, lateral diffusion, film thickness <strong>and</strong><br />
bubblesizebutmostmeasurementsconcentrateontherateoffoamcollapseor,inversely,<br />
the duration of head retention or foam stability. It is obvious that when dispensing beer<br />
either draught or from abottle or can, one can make arough assessment of foam stability<br />
<strong>and</strong> it has been suggested that the views of apanel on this are closer to the consumers'<br />
view of foam behaviour than the tests discussed below. All measurements should be<br />
made in atemperature-controlled room as the collapse of beer foam is very sensitive to<br />
temperature. Scrupulously clean glassware should be used in all the tests, for example,<br />
that washed with 2% trisodium phosphate solution.<br />
Blom (1937) produced foam in atared separating funnel by passing carbon dioxide<br />
through aChamberl<strong>and</strong> c<strong>and</strong>le. The method is therefore applicable to worts <strong>and</strong> aqueous<br />
solutionscontainingfoam-activesubstances.Whenthefunnelisfulloffoam,thebeerisrun<br />
off <strong>and</strong> the funnel weighed. The beer is again run off <strong>and</strong> the residual foam weighed after<br />
intervals of one, two, three, <strong>and</strong> four minutes. Blom observed that the collapse of the foam<br />
was analogous to afirst-order chemical reaction <strong>and</strong> the foam stability (K) was given by:<br />
Kˆ 1<br />
t log<br />
x<br />
…a x†<br />
where tˆtime (min.), aˆinitial weight of the foam, <strong>and</strong> xˆfinal weight of the foam<br />
after time t. By substituting a/2 for xin this equation ,Kcan be calculated from the halflife<br />
of the foam tÝ. Blom found that beers with ahalf-life of 90shave excellent head<br />
retention but values less than 80sindicate poor head retention. Head retention values for<br />
arange of British beers, found using Blom's method, is given in Table 19.23. In Table<br />
19.24 changes in head retention throughout the brewing process are recorded.<br />
If air is passed continuously through a porous membrane into a liquid such as wort or<br />
beer there is a correlation between the volume of air (V) passing in time (t) <strong>and</strong> the<br />
average amount of foam produced ( ) so that:<br />
X ˆ t<br />
P is a measure of the life of a bubble in the foam. Ross <strong>and</strong> Clarke (1939) calculated that<br />
X t<br />
ˆ<br />
2:303 log<br />
b ‡ c<br />
2<br />
where t is the time in seconds of the stationary phase of the head, b is the volume (ml) of<br />
the beer formed in that period, <strong>and</strong> c is the volume of beer given by the residual head <strong>and</strong>:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 19.23 Head retention (Blom) for production beers (Curtis et al, 1963)<br />
X ˆ1:44t1=2<br />
OG t1/2 (s)<br />
Pale ale 1,035 75<br />
Pale ale 1,055 77<br />
Brown ale 1,035 77<br />
Strong ale 1,080 73<br />
Sweet stout 1,046 89<br />
Table 19.24 Changes in head retention (Blom) through the brewing process (Curtis et al.,1963)<br />
OG 1,035 1,055<br />
Wort 115 ±<br />
Beer at rack 86 94<br />
Fined beer 82 88<br />
One day in bottle 78 ±<br />
One week in bottle ± 79<br />
One month in bottle 80 ±<br />
They concluded that was aproperty of the beer foam independent of the method used<br />
to produce it, the dimensions of the container <strong>and</strong> the temperature within alimit of 2ëC.<br />
However, this linear logarithmic decay breaks down when 95% of the foam has returned<br />
to the liquid phase.<br />
The American Society of <strong>Brewing</strong> Chemists have adopted amethod of Helm (1933),<br />
who worked at Carlsberg, in which beer is poured into acylindrical separating funnel<br />
(75mm o.d.) until the foam reaches the 800ml mark. After 30 seconds, all the beer that<br />
has separated is drawn off, the stopwatch started <strong>and</strong> after 200 seconds the beer that has<br />
separated in that period is drawn off (`b' ml) <strong>and</strong> the time `t' noted (225±230s). The<br />
remaining foam is then collapsed, with either isopropyl or butyl alcohol (2ml) delivered<br />
with afine pipette, <strong>and</strong> collected (`c' ml after deducting the volume of the defoaming<br />
agent). The sigma ( )value is then calculated according to Ross <strong>and</strong> Clarke's formula<br />
given above.<br />
The Institute of <strong>Brewing</strong> have adopted amodification of Rudin's (1957) method. The<br />
apparatus (Fig 19.6) consists of ajacketed foam tube (26±28mm dia.) <strong>and</strong>at least 350mm<br />
tallmountedoveraporosity3glasssintereddisc.Degassedbeerisaddedtothe10cmmark<br />
<strong>and</strong>foamedwithcarbondioxidetothe325mmmark.Asthefoamcollapsesthetimetaken<br />
for the foam/beer boundary to traverse the distance between the 50mm <strong>and</strong> the 75mm<br />
marks gives ameasure of the half life of the foam. The logarithmic rectilinear collapse of<br />
foamsformedwithcarbondioxideisfourtimesfasterthanthatforbeersfoamedwithairor<br />
nitrogen so traces of air either in the CO2 used for foaming or introduced into beer by<br />
pouring can cause departures from aregular logarithmic decay. The Institute of <strong>Brewing</strong><br />
alsoapprovethemeasurementoffoamstabilityusingaNIBEMmeter(Klopper,1973)(Fig<br />
19.7).Thefoamisdispensedintoast<strong>and</strong>ardglassoverwhichthemeterheadwithacentral<br />
electrode <strong>and</strong> four shorter needle electrodes is mounted. When the needle electrodes are in<br />
the foam the conductivity between the longer <strong>and</strong> shorter electrodes switches off the<br />
servomotor. As the foam collapses the conductivity is broken the servomotor engages <strong>and</strong><br />
lowers the electrodes until they touch the surface of the foam again After a `wait' period,<br />
the time for the foam to collapse over 10, 20, <strong>and</strong> 30 mm is measured for a beer containing<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Water<br />
outlet<br />
20°C<br />
Side<br />
arm (F)<br />
Trap (D)<br />
Jacketed<br />
foam tube<br />
Water<br />
inlet<br />
Sinter<br />
(b)<br />
Edwards<br />
needle<br />
valve (C)<br />
Flow<br />
indicator<br />
Needle valve (A)<br />
CO2 from cylinder<br />
fitted with reducing<br />
valve<br />
Fig. 19.6 Rudin head retention apparatus (Institute of <strong>Brewing</strong>, Methods of Analysis, 1997).<br />
> 3.4 g/l CO2. For a beer containing < 3.4 g/l CO2 the collapse is measured over 5, 10, <strong>and</strong><br />
15 mm. The IoB Analysis Committee (Sharpe, 1997) found the precision of both the Rudin<br />
<strong>and</strong> the NIBEM methods was independent of the of the foam stability of the sample.<br />
However, the two methods ranked three beers differently, no doubt due to the different<br />
principles involved. A later model, the NIBEM-T meter, has protection against air<br />
movement <strong>and</strong> automatic temperature compensation. It gave better repeatability <strong>and</strong><br />
reproducibility for the determination of foam stability in beer <strong>and</strong> has been accepted by the<br />
EBC Analysis Committee (Ferreira, 2003).<br />
Servomotor to<br />
raise <strong>and</strong> lower<br />
electrodes<br />
120 mm<br />
60 mm<br />
Electrodes<br />
Foam<br />
Beer<br />
Fig. 19.7 NIBEM foam stability apparatus (after Klopper, 1973).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The American Society of <strong>Brewing</strong> Chemists also describe a foam flashing method for<br />
bottled beers. Under a positive pressure of 29 lb CO2 the beer is foamed through a<br />
0.79 mm (0.031 in.) orifice <strong>and</strong> 200 ml of foam are collected. The volume of the beer<br />
collapsed from the foam in 90 seconds is measured (B1) <strong>and</strong> the remaining foam is<br />
collapsed with isopropyl alcohol (2 ml). If the total volume of beer produced from the<br />
collapse of 200 ml of foam, after subtraction of 2 ml of isopropyl alcohol, B2 is then:<br />
Foam Value Units…FVU† ˆ 200…B2 B1†<br />
B2<br />
Several workers prefer to use a st<strong>and</strong>ard orifice since it has been observed that variations<br />
in the porosity of Chamberl<strong>and</strong> c<strong>and</strong>les <strong>and</strong> sintered glass discs can influence the<br />
observed head retention. Ault et al. (1967), Klopper (1973), <strong>and</strong> Jackson <strong>and</strong> Bamforth<br />
(1982) have described methods for measuring cling, lacing or foam adhesion.<br />
19.6.2 Beer components influencing head retention<br />
Pure liquids do not give stable foams <strong>and</strong> many investigations have been carried out to<br />
characterize the surface-active substances in beer responsible for good head retention but<br />
it is obvious, for example, from the use of alcohols to collapse beer foams, that beer<br />
contains both positive <strong>and</strong> negative foam factors. Undoubtedly the most important<br />
positive foam factors in beer are polypeptides. Evans <strong>and</strong> Sheehan (2002) showed that<br />
measurement of beer proteins by the Bradford Coomassie Blue dye binding assay, which<br />
only measures proteins with MW > 5,000, correlated well with Rudin Head Retention<br />
Values. Earlier, Narziss <strong>and</strong> RoÈttger (1973) had found a good correlation between foam<br />
stability (Ross <strong>and</strong> Clarke, 1939) <strong>and</strong> the concentration of nitrogenous material with MW<br />
> 12,000 found by gel filtration. Slack <strong>and</strong> Bamforth (1983) fractionated beer proteins by<br />
hydrophobic interaction chromatography on Octyl-Sepharose CL-4B <strong>and</strong> found that the<br />
most hydrophobic polypeptides were the most foam active.<br />
At least two barley proteins, protein Z <strong>and</strong> lipid transfer protein 1, survive the brewing<br />
process more or less intact. Protein Z (Mr c. 40 kDa), which accounts for 10±25% of the<br />
non-dialyzable protein in beer, can be resolved into two isoforms, Z4 (80%) <strong>and</strong> Z7<br />
(20%). Evans et al. (1999) used ELISA to measure the levels of protein Z4, protein Z7,<br />
BSZ7b, <strong>and</strong> lipid transfer protein 1 (LPT 1) in 25 different malts which were subjected to<br />
pilot or small-scale brewing trials. Regression analyses correlated the foam stability<br />
(Rudin) of the beer with protein (Coomassie blue), protein Z4 (ELISA), free amino<br />
nitrogen (FAN), -glucan, arabinoxylan, <strong>and</strong> viscosity. The levels of protein Z7 <strong>and</strong> LPT<br />
1 were not correlated with the HRV: LTP 1 influences the quantity of foam generated<br />
which is not measured by Rudin's method. By multiple linear regression analysis Evans<br />
et al. (1999) found that the level of malt protein Z4 <strong>and</strong> the wort -glucan level predicts<br />
72% of the variation in foam stability.<br />
Jegou et al. (2000) also found that LTP 1 <strong>and</strong> LTPb in barley grain survived in a<br />
modified form into beer. The modifications, which may influence foam promotion<br />
properties, include glycation with a number of hexose units <strong>and</strong> reduction of disulphide<br />
bonds. Some hordein-derived fragments are also enriched in beer foam (Evans <strong>and</strong><br />
Sheehan, 2002). Varg et al. (1999) identified a 17 kDa protein in barley which is<br />
important for beer foam stability; this protein is partly modified into a more active form<br />
during mashing <strong>and</strong>/or wort boiling.<br />
High molecular weight non-starch polysaccharides such as -glucans <strong>and</strong> arabinoxylans<br />
increase beer viscosity thereby slowing the drainage of liquid from the foam <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
improving its stability. In the trial brews mentioned above (Evans <strong>and</strong> Sheehan, 2002)<br />
viscosity was significantly correlated with HRV <strong>and</strong> with the non-starchy polysaccharides<br />
inbeer. However,high levels of -glucan <strong>and</strong> arabinoxylanare negativelycorrelated<br />
with beer filtration rate so the improvement of foam quality by increasing the level of<br />
non-starchy polysaccharides is not advised. Evans <strong>and</strong> Sheehan (2002) found in atrial<br />
brewthatadditionofBioglucanase <strong>and</strong>Bio-cellulaseduringmashingdegraded morethan<br />
75% of the -glucan <strong>and</strong> reduced the viscosity <strong>and</strong> HRV of the beer but not the lacing<br />
index. In contrast, Lusk et al. (2001) found that no foam enrichment or foam stabilization<br />
occurred with -glucan; it did not collect in the foam <strong>and</strong> treatment with -glucanase did<br />
not alter beer properties.<br />
The hop iso- -acids are concentrated into beer foam. For example, when lager<br />
containing 25±26ppm iso- -acids (50l) was foamed, the collapsed foam (2 l) contained<br />
93±120 ppm iso- -acids (Bishop et al., 1974). Of the individual iso- -acids,<br />
isohumulone <strong>and</strong> isoadhumulone are concentrated into the foam more than isocohumulone.<br />
Similarly, unhopped beer bittered (to 21.0 BU) with isohumulone had abetter head<br />
retention( ˆ132)thanthatbitteredwithisocohumulone( ˆ115)(Difforetal.,1978).<br />
The reduced iso- -acids stabilize foam more than their unsaturated parents. Baker (1990)<br />
added reduced iso- -acid preparations to unhopped beer <strong>and</strong> found that, in particular,<br />
foam collapse times of beers treated with tetrahydroiso- -acids <strong>and</strong> hexahydroiso- -<br />
acids increased significantly with each incremental addition. The foam produced with<br />
hexahydroiso- -acids was unnaturally dense but tetrahydroiso- -acids can be added to<br />
beer to improve the head retention. The amounts that can be added are limited by their<br />
more potent bitterness (Table 8.3) but the addition of 3±5mg/l of tetrahydroiso- -acids<br />
can effectively stabilize beer foam.<br />
Similarly, Weiss et al. (2002), using aNIBEM foam stability meter, found that -iso-<br />
-acids slightly improved the foam stability but the tetrahydroiso- -acids showed amore<br />
dramatic effect. In the presence of iso- -acids metal ions improve beer foaming. Hughes<br />
<strong>and</strong> Simpson (1995) proposed that metals such as Mn 2+ ,Al 3+ ,<strong>and</strong> Ni 2+ cross link the iso-<br />
-acids (Fig 19.8) to strengthen the bubble film. It is well known that polyphenols react<br />
with proteins but despite this they do not appear to have a large effect on foam stability.<br />
In contrast, melanoidins can form stable foams in the absence of protein as they slow the<br />
drainage of liquids from the foam. Finally, the pH of the beer influences foam stability;<br />
the lower the pH the more stable the foam. This may be due to the fact that at lower pH<br />
values the iso- -acids are less ionized <strong>and</strong> more hydrophobic.<br />
Traditionally, the only gas in beer was carbon dioxide but in 1964 Guinness<br />
introduced a draught stout with a mixture of carbon dioxide <strong>and</strong> nitrogen for dispense.<br />
The nitrogen gave the beer its characteristic creamy head since it produces smaller<br />
bubbles than CO2. Afterwards some ales were dispensed in a similar manner. Later, in the<br />
UK, small plastic devices called widgets, were put into beer cans along with liquid<br />
nitrogen. Nitrogen is not very soluble in beer <strong>and</strong> diffuses into the cavity in the widget.<br />
When the can is opened <strong>and</strong> the top pressure released, the nitrogen gas within the widget<br />
at the bottom of the can forces its way through the beer producing considerable amounts<br />
of creamy foam. A warm can may show drastic overfoaming!<br />
Alcohol-free beers usually have unstable foams <strong>and</strong> additions of small amounts of<br />
ethanol (c. 1%) enhances foamability <strong>and</strong> foam stability but this declines at higher<br />
concentrations. Ethanol is weakly surface active but at the concentration in beer (2.3±5.5%,<br />
0.5±1.2 M) it is likely to be challenging foam stability. Lienert (1955) showed that the<br />
straight chain higher alcohols, <strong>and</strong> their acetate esters, all destroy foam, the effect<br />
increasing with the length of the carbon chain. But the major potential foam inhibitors in<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
O<br />
O<br />
HO<br />
R<br />
H<br />
O<br />
O<br />
M 2+<br />
O<br />
O<br />
H<br />
(n – 2) +<br />
Fig. 19.8 Possible structure of iso- -acids chelated to ametal cation (Baxter <strong>and</strong> Hughes,<br />
2001, Fig. 2.2). Reproduced by permission of The Royal Society of Chemistry<br />
beerarelipids (Dickie etal.,2001).Lipidsinbeercanbeeitherintrinsic,derivedfrommalt<br />
<strong>and</strong>yeast lipids,or extrinsic, derived fromdirty glassware, fattyfoodsor cosmetics. Few of<br />
thelipidspresentinthegristarethoughttosurviveintobeersotheintrinsiclipidsarelikely<br />
to be derived from the yeast. For example, asignificant negative correlation was found<br />
between the HRV <strong>and</strong> the free fatty acid content of beer. Bamforth <strong>and</strong> Jackson (1983)<br />
foundthat,inparticular,decanoic<strong>and</strong>dodecanoicacids<strong>and</strong>unsaturatedC 18acidsdisrupted<br />
beerlacing. Monopalmitin was moredisruptiveth<strong>and</strong>ipalmitin <strong>and</strong>tripalmitin (3ppm)had<br />
apositive effect on lacing. Roberts et al. (1978) observed that after the addition of lipid to<br />
beer there was a large drop in the HRV but on stirring the beer the HRV recovers<br />
approaching the original value. This is thought to be due to the presence of lipid binding<br />
proteins in the beer, which have yet to be characterized. Lipid binding proteins have been<br />
characterized in wheat (puroindolines) <strong>and</strong> barley (hordoindolines )(Mr c. 13kDa) but it is<br />
not known whether they survive the brewing process.<br />
19.6.3 Head retention <strong>and</strong> the brewing process<br />
Dilution experiments show that all-malt worts provide an excess of foam-positive<br />
substances so that dilution of the grist with nitrogen-free adjuncts or sugars should not<br />
affect the head retention significantly. The use of unmalted cereals such as wheat flour or<br />
flaked barley in the grist improves head retention in the beer but maize <strong>and</strong> rice have to<br />
be processed before use to reduce the level of lipid. Brewhouse procedures can also<br />
influence the level of lipid in the wort. The last runnings from the mash tun may be rich<br />
inmaltfat.Strainmasters(Nootertuns)givewortswithhighlipidcontentsbutifthemash<br />
worts are recycled the level of lipids is reduced (Chapter 6). Hop-boiling improves head<br />
retention unless so prolonged that all the foam stabilizing proteins are coagulated. The<br />
hop back is more efficient than the hop separator for eliminating lipids.<br />
During the brewing process the largest loss of head retention occurs during<br />
fermentation (Table 19.24). This is due to the loss of foam stabilizing material into the<br />
foam <strong>and</strong> yeast crop <strong>and</strong> to the formation of ethanol, higher alcohols, C6 C12 fatty acids<br />
<strong>and</strong> other negative foam factors. More foam stabilizing factors are lost on top-fermenting<br />
yeasts than bottom-fermenting strainssothe choice ofyeaststrainisimportant. Foam can<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
R<br />
OH<br />
O<br />
O
O<br />
HO<br />
be controlled by the use of enclosed fermentation vessels <strong>and</strong> by recycling part of the<br />
fermenting wort <strong>and</strong> using it to sparge the yeast head <strong>and</strong> foam (Thompson et al., 1965).<br />
Yeast should not be allowed to autolyse in contact with the beer as the fatty acids<br />
liberated will destroy the foam stability. Fining, by means of isinglass, always improves<br />
head retention <strong>and</strong> the finished beer should always be h<strong>and</strong>led gently. Finally, pipes, taps<br />
<strong>and</strong> glassware used in dispense must be kept scrupulously clean <strong>and</strong> well rinsed as traces<br />
of detergent can reduce head retention.<br />
As mentioned above the brewer can improve head retention by choice of malt rich in<br />
protein Z4 <strong>and</strong> by the addition of wheat flour to the grist. Foam quality can also be<br />
improved by the post-fermentation addition of tetrahydroiso- -acids. Propylene glycol<br />
alginate (PGA, Fig. 19.9) is apermitted additive to improve foam stability. It is formed<br />
by partial esterification (80±90%) of alginic acid with propylene oxide. Typical levels of<br />
application are c. 50mg/l.<br />
19.7 Gushing<br />
O<br />
O<br />
O<br />
HO<br />
OH<br />
HO<br />
O<br />
OH O<br />
CO2Na<br />
Gushing, wild, overfoaming or jumping beer, as it is variously called, is an undesirable<br />
quality in packaged beer (Gardner, 1973). Abeer is said to gush when, on releasing the<br />
overpressure, innumerable minute bubbles appear throughout the volume of the beer<br />
which rapidly exp<strong>and</strong> <strong>and</strong> displace the contents of the bottle. In severe cases as much as<br />
three-quarters of the contents may be lost. Outbreaks of gushing may be of two types: (i)<br />
Sporadic or transitory or (ii) Epidemic or serious. Transitory gushing may be related to<br />
minor changes in the production process <strong>and</strong> is usually susceptible to specific cures.<br />
Epidemic outbreaks of gushing may affect several breweries in the same area or those<br />
using acommon source of raw material. Many factors have been implicated in gushing:<br />
since it is overfoaming an excess of the materials involved in head formation may be<br />
responsible.Simon(1998) thinksthat overcarbonation may be responsible for somecases<br />
of gushing but other workers say that, although overcarbonation <strong>and</strong> rough h<strong>and</strong>ling may<br />
stimulate gushing in sensitive beers, it will not induce it in normal beers. As noted above,<br />
beers carbonated with amixture of carbon dioxide <strong>and</strong> nitrogen (<strong>and</strong> awidget) are more<br />
likely to gush than those carbonated with CO2 alone.<br />
Krause (1936) suggested that prolonged shaking beats many microbubbles from the<br />
headspace into the beer <strong>and</strong> these microbubbles attract surface active constituents into<br />
their interfaces. When the gas dissolves the hydrophobic surface active shells may not<br />
dissolve <strong>and</strong> remain to act as nuclei for CO2 evolution. Gardner (1973) has reviewed the<br />
evidence in support of Krause's idea but the nature of the nuclei is still not understood.<br />
The nuclei have not been observed by electron microscopy <strong>and</strong>, by this technique, no<br />
difference could be seen between precipitates collected from gushing <strong>and</strong> non-gushing<br />
O<br />
O<br />
O<br />
O<br />
HO<br />
Fig. 19.9 Idealized structure of propylene glycol alginate (Baxter <strong>and</strong> Hughes, 2001, Fig. 2.5).<br />
Reproduced by permission of the Royal Society of Chemistry.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
HO<br />
OH<br />
O
eers. Turbid beers do not have an increased tendency to gush. A correlation has been<br />
found between surface viscosity <strong>and</strong> gushing. Substances known to promote gushing<br />
increase the viscosity which is reduced when antigushing agents are added. A high<br />
surface viscosity does not cause gushing but appears to correlate with the existence of<br />
stable nuclei.<br />
Transitory outbreaks of gushing have been observed due to the precipitation of<br />
microcrystals of calcium oxalate while other outbreaks have been associated with the<br />
presence of heavy metals, in particular, iron, nickel, tin <strong>and</strong> molybdenum. Such outbreaks<br />
can often be cured by the addition, if allowed, of EDTA to the beer; EDTA forms<br />
complexes with calcium <strong>and</strong> the heavy metals. Iron <strong>and</strong> nickel cause gushing only in the<br />
presence of iso- -acids but cobalt shows little tendency to cause gushing. Indeed, in one<br />
outbreak of gushing the addition of 0.2±1.0 ppm of cobalt dramatically reduced the<br />
incidence of gushing but in another it was without effect. In any case the addition of<br />
cobalt causes toxicity problems with heavy drinkers (Long, 1999).<br />
Severe outbreaks of gushing have been associated with the introduction of new<br />
season's malt especially when this has been made from barley harvested under wet<br />
conditions. Such barleys/malts are often contaminated with moulds especially Fusarium<br />
spp. Both Narziss et al. (1990) <strong>and</strong> Niessen et al. (1992) found that the most common<br />
fungal contaminant of barley, associated with gushing, was F. graminearum. Other<br />
species associated with gushing include F. avenaceum, F. culmorum, F. oxysporum,<br />
Alternaria alternata, <strong>and</strong> species of Penicillium, Mucor <strong>and</strong> Rhizopus Wheat infected<br />
with Microdochium nivale var. major also produces beverages with gushing properties.<br />
Scanning electron spectroscopy showed that with Fusarium species the fungal hyphae<br />
<strong>and</strong> mycelia tended to concentrate in the furrow <strong>and</strong> tip of the grain where they were not<br />
readily apparent to the naked eye.<br />
Similarly, in the USA Schwarz et al. (1996) found that malt samples infected with<br />
Fusarium species produced beers with a propensity to gush <strong>and</strong> that the levels of<br />
deoxynivalenol <strong>and</strong> ergosterol (Fusarium metabolites) strongly correlated with the amount<br />
of gushing. However, it is thought unlikely that deoxynivalenol <strong>and</strong> ergosterol promote<br />
gushing themselves but, rather, a polypeptide produced from the barley or malt by the<br />
micro-organism is the active agent. Obviously it is undesirable to use weathered barley for<br />
malting but, if necessary, the addition of formaldehyde to the steep liquor, if allowed, gives<br />
malt that does not produce a gushing beer. Other successful treatment of wild beer include<br />
the use of absorbents such as kaolin (1,000 g/l), bleaching earth (200 g/l), fullers' earth<br />
(200 g/l), <strong>and</strong> nylon (140 g/l) or to increase the hop rate.<br />
The observation that certain isomerized hop extracts provoked gushing led to a<br />
detailed examination of the gushing potential of individual components. The -acids <strong>and</strong><br />
hulupones suppressed gushing while the iso- -acids <strong>and</strong> the humulinic acids had no<br />
influence on gushing behaviour. The most potent gushing agent found among the hop<br />
compounds was dehydrated humulinic acid [2-isovaleryl-4-(3-methyl-2-butenylidene)<br />
cyclopentane-1, 3-dione] (Laws <strong>and</strong> McGuinness, 1972). This occurred only rarely in<br />
isomerized extracts but 25 ppm provoked gushing in most commercial beers. Moir et al.<br />
(1991) prepared an isomerized extract which produced 38% gushing in beer. The unusual<br />
constituents responsible for the gushing activity were dimers of the iso- -acids formed by<br />
the action of oxygen on the iso- -acids in the presence of iron. Other hop compounds<br />
reported to provoke gushing include the abeo-iso- -acids <strong>and</strong> the hexahydroiso- -acids.<br />
Hop extracts contain variable amounts of fatty acids of which long chain saturated fatty<br />
acids promote gushing while unsaturated acids suppress gushing.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
19.8 References<br />
ANDERSON, I. W., MOLZAHN, S. W., ROBERTS, N. B., BELLIA, J. <strong>and</strong> BIRCHALL, D. (1995) Proc. 25th Congr.<br />
Eur. Brew. Conv. Brussels, 1995, 543.<br />
ANDERSON, R. J., CLAPPERTON, J. F., CRABB, D. <strong>and</strong> HUDSON, J. R. (1975) J. Inst. <strong>Brewing</strong>, 81, 208.<br />
ANON (1960) Which, p. 167.<br />
ANON (1967) Which, p. 366.<br />
APPERSON, K., LEIPER, K. A., MCKEOWN, I. P. <strong>and</strong> BIRCH, D. J. S. (2002) J. Inst. <strong>Brewing</strong>, 108, 193.<br />
AULT, R. G., HUDSON, E. J., LINEHAN, A. J. <strong>and</strong> WOODWARD, J. D. (1967) J. Inst. <strong>Brewing</strong>, 73, 558.<br />
AÈ YRAÈ PAÈ AÈ , T., HOLMBERG, J. <strong>and</strong> SELLMANN-PERSSON, G (1961) Proc. 8th Congr. Eur. Brew. Conv.<br />
Vienna, 1961, 286.<br />
BAKER, C. D. (1989) J. Inst. <strong>Brewing</strong>, 95, 267.<br />
BAKER, C. D. <strong>and</strong> UPPERTON, A. M. (1992) J. Inst. <strong>Brewing</strong>, 98, 461.<br />
BAKER, C. W. (1990) Tech. Quart. MBAA, 27, 33.<br />
BAMFORTH, C. W. (1985) J. Inst. <strong>Brewing</strong>, 91, 370.<br />
BAMFORTH, C. W. (1999) J. Amer. Soc. Brew. Chem., 57, 81.<br />
BAMFORTH, C. W. (2000) J. Inst. <strong>Brewing</strong>, 106, 229.<br />
BAMFORTH, C. W. <strong>and</strong> JACKSON, G. (1983) Proc. 18th Congr. Eur. Brew. Conv. London, 1983, 331.<br />
BAXTER, E. D. <strong>and</strong> HUGHES, P. S. (2001) Beer: Quality, Safety <strong>and</strong> Nutritional Aspects. Cambridge, The<br />
Royal Society of Chemistry pp. xiv + 138.<br />
BECH, L. M., VAAG, P., HEINEMANN, B. <strong>and</strong> BREDDAM, K. (1995) Proc. 25th Congr Eur. Brew. Convn.<br />
Brussels, 1995, 561.<br />
BERNSTEIN, L. <strong>and</strong> LAUFER, L. (1977) J. Amer. Soc. Brew. Chem., 35, 21.<br />
BINNS, F., ENSOR, R. J. <strong>and</strong> MACPHERSON, A. L. (1978) J. Sci. Food Agric., 29, 71.<br />
BISHOP, L. R., WHITEAR, A. L. <strong>and</strong> INMAN, W. R. (1974) J. Inst. <strong>Brewing</strong>, 80, 68.<br />
BLOM, J. (1937) J. Inst. <strong>Brewing</strong>, 43, 251.<br />
BOTTOMLEY, R. A. <strong>and</strong> LINCOLN, G. J. (1958) J. Inst. <strong>Brewing</strong>, 64, 50, 53.<br />
BRENNER, M. W., BERNSTEIN, L., BLICK, S. R. <strong>and</strong> BLENKINSHIP, B. K.. (1965) Proc. A. M. Amer. Soc. Brew.<br />
Chem., 187.<br />
BUCKEE, G. K. (1994) J. Inst. <strong>Brewing</strong>, 100, 57.<br />
BUCKEE, G. K. (1995) Ferment, 8, 357.<br />
BUCKEE, G. K. (1997) J. Inst. <strong>Brewing</strong>, 103, 115.<br />
BUCKEE, G. K. <strong>and</strong> MUNDY, A. P. (1994) J. Inst. <strong>Brewing</strong>, 100, 247.<br />
BUCKEE, G. K., MORRIS, T. M. <strong>and</strong> BAILEY, T. P. (1986) J. Inst. <strong>Brewing</strong>, 92, 475.<br />
CANALES, A. M., DE BANCKS, N. M. <strong>and</strong> GARZA, T. L. I. (1970) Proc. A. M. Amer. Soc. Brew. Chem., 75.<br />
CANAS, B. J., HAVERY, D. C., ROBINSON, L. R., SULLIVAN, M. P., JOE, F. R. <strong>and</strong> DIACHENKO, G. W. (1989) J.<br />
Assoc. Off. Analyt. Chem., 72, 873.<br />
CARRINGTON, R., COLLETT, R. C., DUNKIN, I. R. <strong>and</strong> HALEK, G. (1972) J. Inst. <strong>Brewing</strong>, 78, 243.<br />
CERVERA, M. A., NAVARRO, A., MONTORO, R., CATALA, R. <strong>and</strong> YBAÂ NÄ EZ, N. (1989) J. Assoc. Off. Analyt.<br />
Chem., 72, 282.<br />
CHEN, E. C.-H. <strong>and</strong> VAN GHELUWE, G (1979) J. Amer. Soc. Brew. Chem., 37, 91.<br />
CLARKSON, S. P., ORMROD, I. H. L. <strong>and</strong> SHARPE, F. R. (1995) J. Inst. <strong>Brewing</strong>, 101, 191.<br />
COOTE, N. <strong>and</strong> KIRSOP, B. H. (1974) J. Inst. <strong>Brewing</strong>, 80, 474.<br />
CORNELL, J. (2002) J. Amer. Soc. Brew. Chem., 60, 215.<br />
CURTIS, N. S., CLARK, A. G. <strong>and</strong> OGIE, P. J. (1963) J. Inst. <strong>Brewing</strong>, 69, 30.<br />
DALE, C. J. (1990) Ferment, 3, 217.<br />
DALE, C. J. <strong>and</strong> LYDDIATT, A. (1994) J. Inst. <strong>Brewing</strong>, 100, 173.<br />
DALE, C. J. <strong>and</strong> YOUNG, T. W. (1989a) J. Inst. <strong>Brewing</strong>, 95, 35.<br />
DALE, C. J. <strong>and</strong> YOUNG, T. W. (1989b) J. Inst. <strong>Brewing</strong>, 95, 89.<br />
DICKIE, K. H., CANN, C., NORMAN, E. C., BAMFORTH, C. W. <strong>and</strong> MULLER, R. E. (2001) J. Amer. Soc. Brew.<br />
Chem., 59, 17.<br />
DIFFOR, D., LIKENS, S. T., REHBERGER, A. J. <strong>and</strong> BURKHARDT, R. J. (1978) J. Amer. Soc. Brew. Chem., 36, 63.<br />
DONHAUSER, S., WAGNER, D. <strong>and</strong> JACOB, F. (1987) Monats. fuÈr Brauwissenschaft, 40, 328.<br />
DRAWERT, F. <strong>and</strong> TRESSL, R. (1972) Tech. Quart. MBAA, 9, 72.<br />
DUPIRE, S. (1999) J. Inst. <strong>Brewing</strong>, 105, 265.<br />
ENEVOLDSEN, B. S. <strong>and</strong> SCHMIDT, F. (1974) J. Inst. <strong>Brewing</strong>, 80, 520.<br />
EUROPEAN BREWERY CONVENTION (1999) Monograph 27. Symposium on Beer Foam Quality,<br />
Amsterdam, October 1998. NuÈrnberg: Hans Carl, p. 220.<br />
EVANS, D. E. <strong>and</strong> HEJGAARD, J. (1999) J. Inst. <strong>Brewing</strong>, 105, 159.<br />
EVANS, D. E. <strong>and</strong> SHEEHAN, M. C. (2002) J. Amer. Soc. Brew. Chem., 60, 47.<br />
EVANS, D. E., SHEEHAN, M. C. <strong>and</strong> STEWART, D. C. (1999) J. Inst. <strong>Brewing</strong>, 105, 171.<br />
FARRE, R., GIMENO, M. J. <strong>and</strong> LAGARDA, M. J. (1987) J. Inst. <strong>Brewing</strong>, 93, 394.<br />
FERREIRA, A. A. (2003) J. Inst. <strong>Brewing</strong>, 109, 400.<br />
FISHER, S., HAUSER, G. <strong>and</strong> SOMMER, K (1999) Eur. Brew. Convn. Monograph No. 27 Beer Foam Quality,<br />
Amsterdam, p. 37.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
FOOD LABELLING REGULATIONS, THE (1980) SI 1849: 1980. Schedule 5.<br />
FREESTON, M. J. <strong>and</strong> BAKER, C. D. (1993) J. Inst. <strong>Brewing</strong>, 99, 377.<br />
GARDNER, R. J. (1973) J. Inst. <strong>Brewing</strong>, 79, 275.<br />
GJERTSEN, P. (1955) Proc. 5th Cong, Eur. Brew. Convn., Baden-Baden, p. 37.<br />
GLASER, G. (2002) Modern Brewery Age, 53 (12), 6, 32.<br />
GMELCH, F., MAIER, J., ROSSBAUER, G. ZWACK, F. <strong>and</strong> NAST, D. (1989) Brauwelt, 129 (45), 2156, 2158±<br />
2162.<br />
GREENHOFF, K. <strong>and</strong> WHEELER, R. E. (1981) J. Inst. <strong>Brewing</strong>, 87, 35.<br />
GREENSHIELDS, R. N. (1974) J. Sci. Food Agric., 25, 1307.<br />
HARDING, R. J., NURSTEN, H. E. <strong>and</strong> WREN, J. J. (1977) J. Sci. Food Agric., 28, 225.<br />
HARRIS, G. (1965) J. Inst. <strong>Brewing</strong>, 71, 292.<br />
HARRISON, G. A. F., BYRNE, W. J. <strong>and</strong> COLLINS, E. (1965a) Proc. 10th Congr. Eur. Brew. Convn. Stockholm,<br />
1965, p. 352.<br />
HARRISON, G. A. F., BYRNE, W. J. <strong>and</strong> COLLINS, E. (1965b) J. Inst. <strong>Brewing</strong>, 71, 336.<br />
HEGARTY, P. K., PARSONS, R., BAMFORTH, C. W. <strong>and</strong> MOLZAHN, S. W. (1995) Proc. 25th Congr. Eur. Brew.<br />
Convn. Brussels, 1995, 515.<br />
HEJGAARD, J. <strong>and</strong> KAERSGAARD, P. (1983) J. Inst. <strong>Brewing</strong>, 89, 402.<br />
HELM, E. (1933) Wochensch. Brau., 50, 241.<br />
HM CUSTOMS AND EXCISE (1997) Notice 226. Beer Duty (including Update 1, December 2000)<br />
HUDSON, J. R. (1955) J. Inst. <strong>Brewing</strong>, 61, 127.<br />
HUDSON, J. R. (1969) J. Inst <strong>Brewing</strong>, 75, 164.<br />
HUGHES, P. S. <strong>and</strong> SIMPSON, W. J. (1995) Cerevisia Biotechnol., 20, 35.<br />
HYSERT, D. W., MORRISON, N. M. <strong>and</strong> JAMIESON, A. M. (1979) J. Amer. Soc, Brew. Chem. 37, 30.<br />
ILETT, D. S., BURKE, S. <strong>and</strong> SIMPSON, W. J. (1996) J. Sci. Food Agric., 70, 337.<br />
ISHIBASHI, Y., TERANO, Y., FUKUI, N., HONBOU, N., KAKUI, T., KAWASAKI, S. <strong>and</strong> NAKATANI, K. (1996) J.<br />
Amer. Soc. Brew. Chem., 54, 177.<br />
ISHIHARA, S. (1964) Tests for Colour Blindness, Kanehara, Shuppan, Tokyo, Japan.<br />
JACKSON, G. <strong>and</strong> BAMFORTH, C. W. (1982) J. Inst. <strong>Brewing</strong>, 88, 378.<br />
JEGOU, S., DOULIEZ, J. P., MOLLE, D., BOIVIN, P. <strong>and</strong> MARION, D (2000) J. Agric. Food Chem., 48, 5023.<br />
JOHNSON, B. A. <strong>and</strong> JOHANSSON, C.-G. (1999) J. Inst. <strong>Brewing</strong>, 105, 360.<br />
JURADO, J. (2001a) The Brewer International, 1 (10), 41.<br />
JURADO, J. (2001b) The Brewer International, 1(12), 39.<br />
JURADO, J. (2002a) The Brewer International, 2 (2), 20.<br />
JURADO, J. (2002b) The Brewer International, 2 (4), 30.<br />
JURADO, J. (2002c) The Brewer International, 2 (6), 29.<br />
JURADO, J. (2002d) The Brewer International, 2 (9), 33.<br />
JURADO, J. (2002e) The Brewer International, 2 (12), 21.<br />
KAERSGAARD, P. <strong>and</strong> HEJGAARD, J. (1979) J. Inst. <strong>Brewing</strong>, 85, 103.<br />
KALACÏ , P. <strong>and</strong> KRIÂZÏ EK, M. (2003) J. Inst. <strong>Brewing</strong>, 109, 123.<br />
KIENINGER, H. (1978) Brauwelt, 18, 616.<br />
KIENINGER, H. <strong>and</strong> BIRKOVA, V. (1975) Brauwelt, 115, 1250.<br />
KLOPPER, W. J. (1973) Proc. 14th Congr. Eur. Brew. Convn., Salzburg, p. 363.<br />
KLOPPER, W. J., ANGELINO, S. A. G. F., TUNING, B. <strong>and</strong> VERMEIRE, H. A. (1986) J. Inst. <strong>Brewing</strong>, 92, 225.<br />
KOIKE, K., HASHIMOTO, N., KITAMI, H. <strong>and</strong> OKEDA, K. (1972) Reports of the Research Laboratories of the<br />
Kirin Brewery Co. Ltd., 15, 25.<br />
KONDO, H., YOMO, H., FURUKUBO, S., FUKUI, N., NAKATANI, K. <strong>and</strong> KAWASAKI, Y. (1999) J. Inst. <strong>Brewing</strong>,<br />
105, 293.<br />
KRAUSE, B. (1936) Svenska BryggarefoÈrenings MaÊnadsblad, 51, 221.<br />
LAWS, D. R. J. <strong>and</strong> MCGUINNESS, J. D. (1972) J. Inst. <strong>Brewing</strong>, 78, 302.<br />
LERMUSIEAU, G., BULENS, M. <strong>and</strong> COLLIN, S. (2001) J. Agric. Food Chem., 49, 3867.<br />
LIENERT, H. (1955) Proc. 5th Congr. Eur. Brew. Convn., Baden-Baden, p. 282.<br />
LONG, D. G. (1999) J. Inst. <strong>Brewing</strong>, 105, 79.<br />
LUSK, L. T., GOLSTEIN, H. <strong>and</strong> RYDER, D. (1995) J. Amer. Soc. Brew. Chem., 53, 93.<br />
LUSK, L. T., DUNCOMBE, G. R., KAYE, S. B., NAVARRO, A. <strong>and</strong> RYDER, D. (2001) J. Amer. Soc. Brew. Chem.,<br />
59, 183.<br />
MACKIE, A. E. <strong>and</strong> SLAUGHTER, J. C. (2000) J. Inst. <strong>Brewing</strong>, 106, 209.<br />
MANDL, B. <strong>and</strong> PIENDL, A (1971) Proc. 13th Congr. Eur. Brew. Convn., Estoril, p. 343.<br />
MARTIN, P. A. (1982) J. Inst. <strong>Brewing</strong>, 88, 320.<br />
MCFARLANE, W. D. <strong>and</strong> THOMPSON, K. D. (1964) J. Inst. <strong>Brewing</strong>, 70, 467.<br />
MCMURROUGH, I. (1995) Ferment, 8, 39.<br />
MCMURROUGH, I., MADIGAN, D., DONNELLY, D., HURLEY, J., DOYLE, A. M., HENNIGAN, G., MCNULTY, N. <strong>and</strong><br />
SMYTHE, M. R. (1996) J. Inst. <strong>Brewing</strong>, 102, 327.<br />
MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 151.<br />
MOIR, M. (1994) EBC Monograph XXII. Symposium on Hops, Zoeterwoude, p. 165.<br />
MOIR, M., DOBSON, G. <strong>and</strong> SEATON, J. C. (1991) Proc. 23rd Congr. Eur. Brew. Convn., Lisbon, p. 225.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
MORGAN, K. (1965) J. Inst. <strong>Brewing</strong>, 71, 166.<br />
MORRIS, T. M. (1987) J. Inst. <strong>Brewing</strong>, 93, 13.<br />
MUNDY, A. P. (1991) J. Inst. <strong>Brewing</strong>, 97, 45.<br />
MUNDY, A. P. (1996) J. Inst. <strong>Brewing</strong>, 102, 69.<br />
MUNDY, A. P. <strong>and</strong> BOLEY, N. (1999) J. Inst. <strong>Brewing</strong>, 105, 75.<br />
NARZISS, L. <strong>and</strong> ROÈ TTGER, W. (1973) Brauwissenschaft, 26, 261.<br />
NARZISS, L., MIEDANER, H. <strong>and</strong> BOURJA, T. (1979) Brauwissenschaft, 32, 62.<br />
NARZISS, L., BACK, W., REICHENEDER, E., SIMON, A. <strong>and</strong> GRANDL, R. (1990) Monatsschrift fuÈr<br />
Brauwissenschaft, 43, 296.<br />
NIESSEN, L., DONHAUSER, S., WIEDENEDER, A., GEIGER, E. <strong>and</strong> VOGEL, H. (1992) Brauwelt, 132 (16/17),<br />
702.<br />
OTTER, G. E. <strong>and</strong> TAYLOR, L. (1967) J. Inst. <strong>Brewing</strong>, 73, 570.<br />
OTTER, G. E. <strong>and</strong> TAYLOR, L. (1971) J. Inst. <strong>Brewing</strong>, 77, 467.<br />
OWADES, J. L. <strong>and</strong> JACEVAC, J. (1959) Proc. Ann. Meet. Amer. Soc. Brew. Chem., p. 18.<br />
PALAMAND, S. R. <strong>and</strong> GRIGSBY, J. H. (1974) Brewers Digest, Sept. pp. 58, 90.<br />
PALAMAND, S. R., HARDWICK, W. A. <strong>and</strong> MARKL, K. S. (1969) Proc. Ann. Meet. Amer. Soc. Brew. Chem.,<br />
p. 54.<br />
PAUL, A. A. <strong>and</strong> SOUTHGATE, D. A. T. (eds) (1978) McCance <strong>and</strong> Widowsons' The Composition of Foods.<br />
4th edn HMSO, London.<br />
PEPPARD, T. L. (1985) J. Inst. <strong>Brewing</strong>, 91, 364.<br />
PEPPARD, T. L. <strong>and</strong> HALSEY, S. A. (1981) J. Inst. <strong>Brewing</strong>, 87, 85.<br />
PEPPARD, T. L. <strong>and</strong> HALSEY, S. A. (1982) J. Inst. <strong>Brewing</strong>, 88, 309.<br />
PERPEÁ TE, P. <strong>and</strong> COLLIN, S. (1999) J. Agric. Food Chem., 47, 2374.<br />
PORTNO, A. D. (1966) J. Inst. <strong>Brewing</strong>, 72, 193, 458.<br />
POSTEL, W. (1976) Brauwissenschaft, 29, 39.<br />
POSTEL, W., DRAWERT, F. <strong>and</strong> GUÈ VENC, U. (1972a) Brauwissenschaft, 25, 341.<br />
POSTEL, W., DRAWERT, F. <strong>and</strong> GUÈ VENC, U. (1972b) Brauwissenschaft, 25, 391.<br />
POSTEL, W., DRAWERT, F. <strong>and</strong> GUÈ VENC, U. (1973) Brauwissenschaft, 26, 46.<br />
POSTEL, W., DRAWERT, F. <strong>and</strong> GUÈ VENC, U. (1974) Brauwissenschaft, 27, 11.<br />
POSTEL, W., GORG, A. <strong>and</strong> GUÈ VENC, U. (1976) Brauwissenschaft, 29, 132.<br />
POSTEL, W., GORG, A., DRAWERT, F. <strong>and</strong> GUÈ VENC, U. (1975) Brauwissenschaft, 28, 301.<br />
POSTEL, W., MEIER, B. <strong>and</strong> MARKERT, R. (1983) Monatschrift fuÈr Brauwissenschaft, 36, 360.<br />
PRINS, A. <strong>and</strong> VAN MARLE, J. T. (1999) EBC Monograph 27. Symposium on Beer Foam Quality,<br />
Amsterdam, p. 26.<br />
QURESHI, A. A., BURGER, W. C. <strong>and</strong> PRENTICE, N. (1979) J. Amer. Soc. Brew. Chem. 37, 153.<br />
ROBBERECHT, H., VAN SCHOOR, O. <strong>and</strong> DEELSTRA, H. (1984) J. Food Sci., 49, 300.<br />
ROBERTS, R. T., KEENEY, P. J. <strong>and</strong> WAINWRIGHT, T. (1978) J. Inst. <strong>Brewing</strong>, 84, 9.<br />
RONTELTAP, A., HOLLEMANS, M., BISPERINK, C. G. J. <strong>and</strong> PRINS, A. R. (1991) Tech. Quart. MBAA, 28, 25.<br />
ROSENDAL, I. AND SCHMIDT, F. (1987) J. Inst. <strong>Brewing</strong>, 93, 373.<br />
ROSS, S. <strong>and</strong> CLARKE, C. L. (1939) Wallerstein Labs. Commun., 6, 46.<br />
RUDIN, A. D. (1957) J. Inst. <strong>Brewing</strong>, 62, 506.<br />
SADOSKY, P., SCHWATZ, P. B. <strong>and</strong> HORSLEY, R. D. (2002) J. Amer. Soc. Brew. Chem., 60, 153.<br />
SCHROPP, P., BRUDER, T. <strong>and</strong> FORSTNER, A. (2002) Monatsschrift fuÈr Brauwissenschaft, 55, 212.<br />
SCHWARZ, P. B. <strong>and</strong> HAN, J.-Y. (1995) J. Amer. Soc. Brew. Chem., 53, 157.<br />
SCHWARZ, P. B., BEATTIE, S. <strong>and</strong> CASPER, H. H. (1996) J. Inst. <strong>Brewing</strong>, 102, 93.<br />
SHAH, S. K. (1975) J. Inst. <strong>Brewing</strong>, 81, 293.<br />
SHARPE, F. R. (1997) J. Inst. <strong>Brewing</strong>, 103, 277.<br />
SHARPE, F. R., GARVEY, T. B. <strong>and</strong> PYNE, N. S. (1992) J. Inst. <strong>Brewing</strong>, 98, 321.<br />
SHEEHAN, M. C. <strong>and</strong> SKERRITT, J. H. (1997) J. Inst. <strong>Brewing</strong>, 103, 297.<br />
SHIMIZU, C., NAKAMURA, Y., MIYAI, K., ARAKI, S., TAKASHIO, M. <strong>and</strong> SHINOTSUKA, K. (2001) J. Amer. Soc.<br />
Brew. Chem., 59, 51.<br />
SIEBERT, K. J. (1999) J. Agric. Food Chem., 47, 353.<br />
SIEBERT, K. J. <strong>and</strong> LYNN, P. Y. (1997) J. Amer. Soc. Brew. Chem., 55, 73.<br />
SIEBERT, K. J. <strong>and</strong> LYNN, P. Y. (1998) J. Amer. Soc. Brew. Chem., 56, 24.<br />
SIEBERT, K. J., TROUKHANOVA, N. V. <strong>and</strong> LYNN, P. Y. (1996) J. Agric. Food Chem., 44, 80.<br />
SILBEREISEN, K. <strong>and</strong> BIELIG, K. (1961) Proc. 8th Congr. Eur. Brew. Convn., Vienna, p. 421.<br />
SIMON, R. (1998) Brauwelt, 138, 1244.<br />
SIMPSON, W. J. <strong>and</strong> HUGHES, P. S. (1994) Cerevisia Biotechnol., 19 (3), 39.<br />
SINCLAIR, A., HALL, R. D., THORBURN-BURNS, D. <strong>and</strong> HAYES, W. P. (1970) J. Sci. Food Agric., 21, 468.<br />
SLACK, P. T. <strong>and</strong> BAMFORTH, C. W. (1983) J. Inst. <strong>Brewing</strong>, 89, 397.<br />
SLAUGHTER, J .C. <strong>and</strong> UVGARD, A. R. A. (1971) J. Inst. <strong>Brewing</strong>, 77, 446.<br />
SLAUGHTER, J. C. <strong>and</strong> UVGARD, A. R. A. (1972) J. Inst. <strong>Brewing</strong>, 78, 322.<br />
SMEDLEY, S. M. (1992) J. Inst. <strong>Brewing</strong>, 98, 497.<br />
SMEDLEY, S. M. (1995) J. Inst. <strong>Brewing</strong>, 101, 195.<br />
SéRENSEN, S. B. <strong>and</strong> OTTESEN, M. (1978) Carlsberg Res. Commun., 43, 133.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
THALAKER, R. <strong>and</strong> KALTWASSER, I. (1978) Monatsschrift fuÈr Brauerei, 31, 20.<br />
THOMPSON, C. C., CURTIS, N. S., GOUGH, P. E. <strong>and</strong> RALPHM D. J. (1965) Proc. 10th Congr. Eur. Brew, Conv.,<br />
Stockholm, p. 305.<br />
THORNE, R. S. W. <strong>and</strong> NANNESTED, I. (1960) J. Inst. <strong>Brewing</strong>, 66, 388.<br />
THORPE, E. <strong>and</strong> BROWN, H. T. (1914) J. Inst. <strong>Brewing</strong>, 20, 569.<br />
TRESSL, R., KOSSA, T., RENNER, R. <strong>and</strong> KOÈ PPLER, H. (1975) Monatsschrift fuÈr Brauerei, 28, 109.<br />
TRESSL, R., RENNER, R., KOSSA, T. <strong>and</strong> KOÈ PPLER, H. (1977) Proc. 16th Congr. Eur. Brew. Convn.,<br />
Amsterdam, p. 693.<br />
TRESSL, R., FRIESE, L., FENDSACK, F. <strong>and</strong> KOÈ PPLER, H. (1978) J. Agric. Food Chem., 26, 1422.<br />
VAAG, P., BECH, L. M., CAMERON-MILLS, V. <strong>and</strong> SVENDSEN, I. (1999) Proc. 27th Congr. Eur. Brew. Convn.,<br />
Cannes, p. 157.<br />
VAN OEVELEN, D., DE L'ESCAILLE, F. <strong>and</strong> VERACHTERT, H. (1976) J. Inst. <strong>Brewing</strong>, 82, 322.<br />
WARNAKULASURIYA, S., HARRIS, C., GELBIER, S., KEATING, J. <strong>and</strong> PETERS, T. (2002) Clin. Chim. Acta, 320,<br />
1.<br />
WEISS, A., SCHOÈ NBERGER, CH., MITTER, W., BIENDL, M., BACK, W. <strong>and</strong> KROTTENTHALER, M. (2002) J. Inst.<br />
<strong>Brewing</strong>, 108, 236.<br />
WHITTLE, N., ELDRIDGE, H., BARTLEY, J. <strong>and</strong> ORGAN, G. (1999) J. Inst. <strong>Brewing</strong>, 105, 89.<br />
WILLIAMS, K. M., FOX, P. <strong>and</strong> MARSHALL, T. (1995) J. Inst. <strong>Brewing</strong>, 101, 365.<br />
YANG, J.-I. <strong>and</strong> SIEBERT, K. J. (2001) J. Amer. Soc. Brew. Chem., 59, 172.<br />
YBAÂ NÄ EZ, N., NAVARRO, A. <strong>and</strong> MONTORO, R. (1989) J. Inst. <strong>Brewing</strong>, 95, 257.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
20<br />
Beer flavour <strong>and</strong> sensory assessment<br />
20.1 Introduction<br />
The final arbiter of beer quality is the palate of the consumer <strong>and</strong> this can show wide<br />
variation between individuals, between geographical areas, <strong>and</strong> even from one time to<br />
another. Quality is defined as `degree of excellence, relative nature, or kind, or character'<br />
<strong>and</strong> accordingly the brewer refers to the many varieties of ale, beer, stout, <strong>and</strong> lager<br />
which he brews to satisfy varied dem<strong>and</strong> as different qualities. When the customer has<br />
chosen the quality he wishes to drink he dem<strong>and</strong>s that his beverage shall have the `degree<br />
of excellence' which he expects <strong>and</strong> this shall not change from day to day. Much of the<br />
brewers' art is therefore concerned with quality control, with producing a constant<br />
product from variable raw materials by a biological process (Hough, 1990). Today the<br />
concept of quality assurance is more important whereby each stage is monitored before<br />
the next is allowed <strong>and</strong> the raw materials <strong>and</strong> any additives can be traced back to their<br />
source (see EBC Monograph No. 26 Symposium on Quality Issues & HACCP (Hazard<br />
Analysis Critical Control Points), 1997).<br />
The enjoyment of a glass of beer may be received by many senses. Smythe et al.<br />
(2002) have shown the impact of the appearance of beer on its perception <strong>and</strong> the<br />
parameters discussed in the last chapter: alcoholic content, nutritive value, colour,<br />
clarity, usually, the absence of haze, the formation <strong>and</strong> retention of a good head of<br />
foam, <strong>and</strong> the absence of gushing, all contribute to the enjoyment but it is the flavour,<br />
the taste <strong>and</strong> aroma, which really determine the acceptability <strong>and</strong> drinkability of the<br />
beer. Originally, it was the palate of the head brewer that decided if the beer was<br />
acceptable but later this responsibility was transferred to a tasting panel <strong>and</strong> as tasting<br />
panels became more sophisticated the <strong>science</strong> of sensory analysis came into being<br />
(Amerine et al., 1965). Analytica-EBC <strong>and</strong> the ASBC have agreed International<br />
methods for sensory analysis <strong>and</strong> the Institute of <strong>Brewing</strong> have published a Sensory<br />
Analysis Manual (Institute of <strong>Brewing</strong>, 1995). More recent advances have involved the<br />
use of electronic `noses' or sensors (Torline et al., 1999; Given <strong>and</strong> Parades, 2002).<br />
Doty (1995) has edited a H<strong>and</strong>book of Olfaction <strong>and</strong> Gustation <strong>and</strong> Acree <strong>and</strong> Teranishi<br />
have edited a h<strong>and</strong>book of Flavor Science (1993).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
20.2 Flavour ± taste <strong>and</strong> odour<br />
According to the EBC-ASBC definitions, flavour is the combination of olfactory <strong>and</strong><br />
gustatory attributes perceived during tasting, including tactile, thermal, pain <strong>and</strong> kinesthetic<br />
effects. Kinesthetic sense is the deep pressure sense, or proprioception. Somesthetic sense is<br />
the tactile sense, or skin-feel, both are part of the sense of touch. Tactile senses refer more<br />
to solid foodstuffs but in beer are related to what is called `palate fullness', `body', or<br />
`mouth-feel' (Langstaff <strong>and</strong> Lewis, 1993). EBC/ASBC define mouth-feel as the tactile<br />
sensations perceived at the lining of the mouth, including the tongue <strong>and</strong> teeth. With regard<br />
to beer, taste <strong>and</strong> odour are the most important aspects of flavour.<br />
Taste is the sensory attribute resulting from stimulation of the gustatory receptors in<br />
the oral cavity by certain soluble substances. Two American Chemical Society symposia<br />
have discussed taste chemistry (Boudreau, 1979; Given <strong>and</strong> Parades, 2002) <strong>and</strong> Breslin<br />
(2001) has provided a review. Two types of chemicoreceptor are recognized; free nerve<br />
endings, which occur throughout the oral cavity, <strong>and</strong> taste buds. The free nerve endings<br />
possess no recognizable receptors <strong>and</strong> are responsible for the perception of pungency <strong>and</strong><br />
astringency. Taste buds are neural complexes of 25±50 specialized cells which occur in<br />
localized areas of the oral cavity. On the tongue they occur on protuberances called<br />
papillae. Four types of papillae are recognized: the filiform papillae have no taste buds<br />
<strong>and</strong> the foliate papillae, which occur in folds on the sides of the back of the tongue, are<br />
not well developed in man. More important are the 13±400 mushroom-like fungiform<br />
papillae on the tip <strong>and</strong> sides of the tongue <strong>and</strong> the 6±15 large (circum)vallate papillae at<br />
the back of the tongue (Fig. 20.1) (See Amerine et al., 1965, for microphotographs). Taste<br />
buds are pear shaped (40±70 m) connected to the oral cavity via a narrow pore (2 m).<br />
At the top of the taste bud microvilli (0.1±0.2 m diam. 1±2 m long) are situated in the<br />
pore <strong>and</strong> these are probably the first point of contact with the tastants. The taste stimuli<br />
apparently do not penetrate the receptor membrane but interact at the outer surface.<br />
Taste buds<br />
Vallate papillae<br />
Foliate papillae<br />
Fungiform papillae<br />
Pore<br />
Nerve<br />
Sensory<br />
cells<br />
Fig. 20.1 Location of some oral chemosensory receptor systems. Taste buds (schematic, upper<br />
right) are found on specialized papillae on the tongue <strong>and</strong> scattered on the palate <strong>and</strong> posterior<br />
oral structures. Free nerve endings are found on all oral surfaces (After Boudreau, 1979).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Bitter<br />
Sour Salt Sweet<br />
Fig. 20.2 Areas of the human tongue where the four tastes are most easily sensed. All four<br />
tastes are perceived, but less readily, over the central area.<br />
The taste cells are secondary receptor cells as they have no axions of their own but are<br />
connected by basal synapses to taste fibres running to the central nervous system. Three<br />
nerves are involved; the nervus glossopharyngeus (IX) for the back third of the tongue<br />
including the circumvallate papillae, the nervus vagus (X) for the throat <strong>and</strong> the larynx,<br />
<strong>and</strong> the chorda tympani (CT) part of the nervus facialus (VII) for the front two-thirds of<br />
the tongue with the fungiform papillae. It is estimated that the circumvallate papillae<br />
contain 1,000±1,500 taste buds <strong>and</strong> the fungiform papillae 300±400 (Van der Heijden,<br />
1993).<br />
Althoughthefourbasictastes,sweet,sour,salt,<strong>and</strong>bitterareperceivedthroughoutthe<br />
oral cavity, they are perceived more strongly in specific areas (Fig. 20.2). Inspection of<br />
this figure shows that the bitter taste of beer can only be evaluated satisfactorily if the<br />
beerisswallowed<strong>and</strong>allowed toflow overthecircumvallatetastebudsatthebackofthe<br />
tongue. The sweet taste, perceived by taste buds on the fungiform papillae, has received<br />
the most study (Van der Heijden, 1993). The transduction of the sweet taste appears to<br />
involve specific membrane receptors <strong>and</strong> Teeter <strong>and</strong> Gold (1988) have proposed the<br />
transduction pathway shown in Fig. 20.3. Akabas et al. (1988) concluded that<br />
transduction of bitter taste may occur via a receptor-second messenger mechanism<br />
leading to neurotransmitter release <strong>and</strong> may not involve depolarization-mediated calcium<br />
entry. The bitter principle denatonium chloride (Fig. 20.4) is apotent blocker of outward<br />
potassium currents in taste cells <strong>and</strong> causes asecond messenger release of Ca 2+ from<br />
intracellular stores in asubset of taste cells.<br />
Many substances are known to taste bitter (Rouseff, 1990) <strong>and</strong> in view of their diverse<br />
structures (Fig. 20.4) Delwiche et al. (2001) have proposed there must be multiple<br />
receptor/transduction mechanisms. They found that the ratings <strong>and</strong> rankings of 26<br />
subjects placed bitter substances in two general clusters: (i) urea, phenylalanine,<br />
tryptophan, <strong>and</strong> epicatechin, <strong>and</strong> (ii) quinine (20.1), caffeine (20.2), sucrose octa-acetate<br />
(20.3), denatonium benzoate (20.5), tetralone Õ (tetrahydroiso- -acids, 8.47) <strong>and</strong><br />
magnesium sulphate. Neither of these groups included propylthiouracil (PROP) (20.4)<br />
to which tasters show wide variations in sensitivity. Approximately 25% of tasters find<br />
this antithyroid drug extremely bitter (`supertasters'), 50% find it bitter (`tasters') but<br />
25%are non-tasters. It is thought that `supertasters' perceive all tastes, not just bitterness,<br />
more intensely. In `supertasters' the fungiform papillae on the tongue are denser (>35/<br />
38.5mm 2 )comparedwithanaverageof15±35/38.5mm 2 ;non-tastershave
Chemical stimulus<br />
activates GTP-binding protein<br />
stimulates adenylate cyclase<br />
increases intracellular cAMP (or cGMP)<br />
closes apical K + channels<br />
depolarizes taste cell<br />
opens voltage-dependent Ca 2+ channels<br />
allows influx of Ca 2+ near synapses<br />
causes neurotransmitter release at synapses<br />
Fig. 20.3 Transduction pathway for the sweet taste (Teeter <strong>and</strong> Gold, 1988).<br />
CH3O<br />
N<br />
CH(OH)<br />
CH2OAc<br />
AcO·CH2<br />
O<br />
O<br />
OAc<br />
AcO<br />
OAc<br />
O<br />
OAc<br />
N<br />
CH<br />
AcO<br />
CH2OAc<br />
CH3<br />
CH3<br />
CH2<br />
C2H5<br />
C2H5<br />
H3C<br />
O<br />
Pr n<br />
N<br />
O<br />
N<br />
CH3<br />
(20.1) Quinine (20.2) Caffeine<br />
H<br />
N<br />
O<br />
CH3<br />
(20.3) Sucrose octaacetate (20.4) Propylthiouracil<br />
NH·CO·CH2·N + ·CH2<br />
(20.5) Denatonium benzoate<br />
Fig. 20.4 Bitter compounds.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
N<br />
NH<br />
N<br />
S<br />
C6H5·CO2 –
men (Goode, 2003). Delwiche et al. (2001) found that although PROP sensitive <strong>and</strong><br />
PROP insensitive tasters rated the above chemicals differently, they ranked them in a<br />
similar order. Montmayeur <strong>and</strong> Matsunami (2002) have discussed receptors for bitter <strong>and</strong><br />
sweet tastes. The concentration-response curve for most tastant-receptor reactions is<br />
sigmoid in shape (Breslin, 2001).<br />
Other basic tastes include salty which is produced by relatively high concentrations of<br />
inorganic ions, in particular Na + , K + , <strong>and</strong> Li + , on the fungiform papillae but it is seldom a<br />
dominant taste in beer. The sour taste is evoked by various Bronsted acids mainly on the<br />
foliate papillae on the sides of the tongue. In most beers it is regarded as an off-flavour but is<br />
an important character in some Belgian beers produced by spontaneous fermentation.<br />
Pleasant is associated with the small fibre geniculate ganglion system <strong>and</strong> is evoked by<br />
lactones <strong>and</strong> similar carbon-oxygen compounds. `Umani' is a Japanese word (meaning<br />
`deliciousness') used to describe the sensation elicited by the amino acid monosodium<br />
glutamate <strong>and</strong> the nucleotides sodium inosinate <strong>and</strong> sodium guanylate. The metallic<br />
sensation is produced by certain salts such as silver nitrate <strong>and</strong> by oct-1-en-3-one. As<br />
mentioned above pungency <strong>and</strong> astringency are sensations produced at the free nerve<br />
endings.<br />
Odour is more complicated than taste (see, for example, Ohloff, 1994). Orthonasal<br />
olfaction occurs when an odour is sniffed through the nostrils into the nasal cavity where<br />
the receptors are located on the olfactory epithelium in the upper respiratory passages<br />
(Fig. 20.5). In man the olfactory epithelium occupies 2±4 cm 2 <strong>and</strong> contains about 9<br />
million neurons; it is more extensive in other animals. The axions from these cells, many<br />
of which cannot be seen with a light microscope, are grouped together in bundles <strong>and</strong><br />
pass through the cribiform plate into the olfactory bulb, where they terminate in small<br />
bodies known as glomeruli. From the glomeruli, mitral cells pass directly into the<br />
olfactory lobe of the brain. No other stimuli are received by the brain in such a direct<br />
manner. Retronasal olfaction occurs when odours released during eating or drinking are<br />
forced behind the palate into the nasal cavity. However, it is likely that the two forms of<br />
olfactory input are analysed in different parts of the brain. In beer, <strong>and</strong> other beverages<br />
<strong>and</strong> foodstuffs, the strength of the odour impression is partly governed by the volatility of<br />
the molecules from water, i.e., by the air-water partition coefficient.<br />
Frontal sinus<br />
Openings to:<br />
frontal sinus<br />
maxillary sinus<br />
naso-lacrimal duct<br />
Air passages<br />
OLFACTORY<br />
BULB<br />
Middle<br />
turbinate<br />
Inferior<br />
turbinate<br />
OLFACTORY<br />
CENTRES<br />
INCUS<br />
Superior<br />
turbinate<br />
Lined with<br />
COLUMNAR CILIATED<br />
EPITHELIUM<br />
Opening of<br />
eustachian tube<br />
To respiratory<br />
system<br />
Fig. 20.5 Vertical section of the nasal region of the head.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.1 Olfactometric properties of eight primary odourants (Amoore, 1991)<br />
It is thought that the odour perceived is made up from a number of primary odours. Many of<br />
these primary odours have been detected from persons with specific anosmias, `smell<br />
blindness' which is the olfactory analogue of colour blindness (Amoore, 1991). In the<br />
examples studied so far 3±47% of the population have shown specific anosmia for some<br />
compound. Conversely, some people exhibit hyperosmia, when their sensitivity to certain<br />
odours may be 1,000-fold greater than normal. Thirty-six per cent of a population had a<br />
specific anosmia for the malty isobutyraldehyde. These people could still detect<br />
isobutyraldehyde after 16 twofold dilution steps whereas the normal population could detect<br />
this compound after 24 dilution steps (a 500-fold deficit of sensitivity for the anosmics). When<br />
the same panels investigated isobutyl alcohol there was some overlap between the panels <strong>and</strong><br />
the anosmic defect was only 4.1 steps In contrast with isobutyl isobutyrate, the thresholds of<br />
the two groups overlapped. Eight primary odorants, which have been examined in this way, are<br />
listed in Table 20.1 but 76 compounds have been observed for which specific anosmia has been<br />
reported. Several systems of odour classification have been proposed containing 4±44 classes<br />
<strong>and</strong> these have been reviewed by Amoore (1991) together with specific anosmia analyses.<br />
Goodenough (1998) has reviewed the molecular biology of olfactory perception.<br />
Taste <strong>and</strong> odour can be perceived separately but more often than not the two senses are<br />
integrated to produce the sensation of flavour. The measurement of taste, odour, or<br />
flavour intensity is the subject of at least two different approaches which use different<br />
mathematics (Meilgaard, 1975). Those working with strong flavours are concerned with<br />
suprathreshold effects <strong>and</strong> describe the perceived intensity R as a power factor n of the<br />
concentration S so that:<br />
R ˆ constant S n<br />
For example, with butanol in air the equation becomes:<br />
R ˆ 0:261 S 0:66<br />
On the other h<strong>and</strong> those working with more delicate flavours such as food, vegetables,<br />
whisky, wine <strong>and</strong> beer have assumed that the perceived intensity R is proportional to the<br />
concentration S <strong>and</strong> inversely proportional to the threshold concentration T so that:<br />
R ˆ …constant† S=T<br />
Normal threshold<br />
Primary odorant Primary In air l/l In water Anosmic Anosmic<br />
odour (v/v) mg/l (w/v) occurrence defect<br />
(%) factor<br />
Isovaleric acid Sweaty 0.0010 0.12 3 42<br />
1-Pyrroline Spermous 0.0018 0.020 16 39<br />
Trimethylamine Fishy 0.0010 0.00047 6 830<br />
Isobutyraldehyde Malty 0.0050 0.0018 36 340<br />
5 -Androst-16-en-3-one Urinous 0.00019 0.00018 47 770<br />
!-Pentadecalactone Musky 0.018 0.0018 12 13<br />
l-Carvone Minty 0.0056 0.04 8 13<br />
1,8-Cineole Camphorous 0.011 0.020 33 56<br />
The power factor n is assumed to be 1.00 <strong>and</strong> the constant is often omitted so that R is<br />
measured as S/T. This ratio has been given different names including `flavour units' (see<br />
p. 285) (Meilgaard, 1975). The power function n in the equation:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
R ˆ …S=T† n<br />
has been estimated for ethanol (n ˆ 1.54), dimethyl sulphide (n ˆ 1.12), diacetyl (n ˆ<br />
0.89) <strong>and</strong> isoamyl acetate (n ˆ 0.82) (Meilgaard <strong>and</strong> Reid, 1979). Thus, except for<br />
ethanol, the error in assuming n ˆ 1.00 is less than 12%.<br />
The EBC/ASBC define four different types of threshold <strong>and</strong> the term should not be<br />
used without qualification:<br />
1. Detection threshold (stimulus threshold, absolute threshold). The minimum value<br />
of a sensory stimulus needed to give rise to a sensation.<br />
2. Difference threshold (just noticeable difference). Value of the smallest perceptable<br />
change in the physical intensity of a stimulus.<br />
3. Recognition threshold. The minimum value of a sensory stimulus permitting<br />
identification of the sensation received.<br />
4. Terminal threshold. Maximum value of a sensory stimulus permitting identification<br />
of the sensation perceived.<br />
Usually detection thresholds are lower than recognition thresholds but the literature does<br />
not always indicate which is being measured especially when the assessors know the<br />
identity of the test substance. When substances are added to beer we are concerned with<br />
difference thresholds but many of the methods used for estimation involve recognition of<br />
the odd sample. In <strong>practice</strong>, with trained assessors the difference between difference <strong>and</strong><br />
recognition thresholds is negligible (Brown et al., 1978). Threshold values are subject to<br />
considerable biological variation <strong>and</strong> it is therefore desirable to include statistical limits<br />
in any estimation. Criteria which have been used in estimating thresholds include (Brown<br />
et al., 1978):<br />
1. The concentration that can just be perceived by 50% of the population.<br />
2. The geometric mean of the individual thresholds (maximum likelihood threshold).<br />
3. The lowest concentration that can be detected with a ** statistical significance (P ˆ<br />
0.01).<br />
4. The concentration producing 50% correct choices (ASTM). Since in paired sample<br />
tests 50% correct choices can occur by chance alone, this may be amended to:<br />
5. That concentration producing a frequency of 50% correct above chance.<br />
Many factors influence the measurement of thresholds. For example, the influence of<br />
temperature on taste is not uniform <strong>and</strong> the buffering action of saliva (pH 7.0) may<br />
influence perception. The mode of presentation of the sample is also important; thus the<br />
average sensitivity threshold for sodium chloride varied from 0.135% for three drops<br />
placed on the tongue to 0.047% for 10 ml swallowed <strong>and</strong> 0.016% when unlimited<br />
amounts of the salt solution <strong>and</strong> distilled water could be compared (Richter <strong>and</strong> Maclean,<br />
1939). Similarly, repeated <strong>practice</strong> appears to lower the threshold at which tastes can be<br />
perceived. The following thresholds (percentages) were for the first <strong>and</strong> sixth<br />
determinations: sucrose, 0.753, 0.274; caffeine, 0.0272, 0.0078; citric acid, 0.0223,<br />
0.00096; <strong>and</strong> sodium chloride, 0.123, 0.047 (Pangborn, 1959).<br />
There can be large differences in sensitivity from person to person. For the addition of<br />
dimethyl sulphide to beer, tasted by an international panel of 44 persons, 37 panellists<br />
(85%) had thresholds in the range of 12±87 g/l, but the remaining seven panellists were<br />
much less sensitive <strong>and</strong> had thresholds in the range 150±2000 g/l (Brown et al., 1978).<br />
Similar results were obtained with diacetyl; with a panel of 16 assessors the geometric<br />
mean threshold was 0.080 mg/l but one assessor had a threshold of 2.26 mg/l (Meilgaard<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.2 Detection thresholds for various common taste substances (after Breslin, 2001)<br />
Sensation Compound [M] mg/l<br />
Bitter Caffeine<br />
Magnesium Sulphate<br />
Quinine hydrochloride<br />
Sucrose octa-acetate<br />
Urea<br />
PROP (taster)<br />
PROP (nontaster)<br />
Salty NH4Cl<br />
CaCl2<br />
LiCl<br />
NaCl<br />
KCl<br />
Monosodium glutamate<br />
L-Arginine<br />
L-Glutamine<br />
Sweet Aspartame<br />
Fructose<br />
Glucose<br />
Glycine<br />
Saccharin Na<br />
Sucrose<br />
Sour Acetic acid<br />
Citric acid<br />
Hydrochloric acid<br />
Malic acid<br />
Tartaric acid<br />
PROP = Propylthiouracil<br />
5 10 4<br />
3.85 10 4<br />
1.4 10 6<br />
3.58 10 6<br />
(1.07±1.72) 10 2<br />
2 x 10 5<br />
6 10 4<br />
8.39 10 4<br />
8 10 6<br />
(0.9±4) 10 2<br />
1.02 10 3<br />
(6.31±6.49) 10 3<br />
5 10 4<br />
1.23 10 3<br />
9.77 10 3<br />
(1.76±2.08) 10 5<br />
8.9 10 4<br />
7.33 10 3<br />
3.09 10 2<br />
(8.58±10.1) 10 6<br />
6.5 10 4<br />
(1.07±1.12) 10 4<br />
7 10 5<br />
1.6 10 4<br />
7.3 10 5<br />
4.78 10 5<br />
97<br />
46.35<br />
0.505<br />
2.43<br />
642±1033<br />
4.04<br />
121.2<br />
44.89<br />
0.888<br />
381±1696<br />
59.62<br />
470±484<br />
84.5<br />
214.3<br />
1428<br />
5.17±6.12<br />
160<br />
1321<br />
2317<br />
1.76±2.07<br />
222.5<br />
6.43±6.73<br />
13.45<br />
5.84<br />
9.79<br />
7.17<br />
<strong>and</strong> Reid, 1979). Thus, when measuring thresholds it is desirable to have apanel of at<br />
least 25 persons <strong>and</strong> to calculate individual thresholds so that insensitive persons do not<br />
over-influence the group result (Meilgaard <strong>and</strong> Reid, 1979; Brown et al., 1978).<br />
Analytica-EBC/ASBC give methods to determine the threshold of an added substance by<br />
the ascending method of limits (see later).<br />
Detection thresholds for representative compounds are collected in Table 20.2. Other<br />
values are given byMeilgaard (1975) <strong>and</strong> some can be found inthe Tables inChapter19.<br />
Quinine (20.1) is often considered the st<strong>and</strong>ard bitter taste <strong>and</strong> is the most bitter<br />
compound listed in Table 20.1 but denatonium salts are more bitter. Quinine tonic water<br />
contains 56mg/l (0.5 grain/pint) of quinine sulphate together with sugar (45g/l) <strong>and</strong>/or<br />
permitted artificial sweetening agents. Caffeine (20.2) is much less bitter <strong>and</strong> is found in<br />
proprietary soft drinks such as Coca-Cola <strong>and</strong> Lucozade as well as tea <strong>and</strong> coffee.<br />
Magnesium sulphate has similar bitterness to caffeine <strong>and</strong> it is noteworthy that with<br />
propylthiouracil (20.4) the threshold for tasters is 30 times less than that for non-tasters.<br />
Sodiumchloridehasthecharacteristic saltytasteabove0.05 M,butweakersolutions taste<br />
sweet. Potassium chloride also tastes sweet in dilute solutions <strong>and</strong> salty above 0.05 Mbut<br />
between 0.02 M<strong>and</strong> 0.03 Monly bitterness is perceived <strong>and</strong> this note persists at higher<br />
concentration. Calcium chloride has the lowest threshold amongst the salts quoted. The<br />
artificial sweeteners aspartame <strong>and</strong> saccharin have much lower thresholds than the<br />
natural sugars. The relative sweetness of anumber of other sugars is given in Table 20.3.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.3 Relative sweetness of sugars (Nieman, 1960)<br />
Lactose 39<br />
Maltose 46<br />
D-Xylose 67<br />
, -D-Glucose 69<br />
Glycerol 79<br />
Invert sugar 95<br />
Sucrose 100<br />
Fructose 114<br />
Calcium cyclamate 3,380<br />
Saccharin 30,000<br />
With mineral acids, such as hydrochloric acid, the sourness is proportional to the<br />
hydrogen ion concentration but this is not the case with organic acids that are largely<br />
undissociated. At equimolecular concentration, hydrochloric acid tastes more sour than<br />
acetic acid but at the same pH, acetic acid tastes sourer than the mineral acid.<br />
Presumably,astheH + ofthedissociatedaceticacidreactswiththetastereceptor,someof<br />
the undissociated acid ionizes to restore the equilibrium.<br />
Reviewing the primary tastes inbeeritssourness will bemeasured either asitspH(3.8±<br />
4.7) or as its titratable acidity. The level of acetic acid reported for normal beers (57±<br />
145ppm) is below the taste threshold <strong>and</strong> the same is true for lactic acid (Table 19.7).<br />
Infectionofbeerwithmicro-organismssuchasAcetobacterspp.or Lactobacillusspp.may<br />
reduce the pH <strong>and</strong> produce asour taste. Thus with lambic <strong>and</strong> gueuze beer the pH may<br />
drop to 3.2. The level of acetic acid in these beers is 2.6±6.9 times the taste threshold <strong>and</strong><br />
the level of lactic acid is 5.8±8.6 times the taste threshold (Van Oevelen et al., 1976).<br />
Comparison of the level of the inorganic constituents of beer (Table 19.2) with the<br />
thresholds give in Table 20.2 suggests that the thresholds of potassium chloride, sodium<br />
chloride,calciumchloride <strong>and</strong>magnesium sulphate couldbeexceeded, butbeers are rarely<br />
classed as salty. Similarly, the levels of fructose <strong>and</strong> glucose in the primed beers 2±4 in<br />
Table 19.5 exceed the taste thresholds <strong>and</strong> the level of glucose in the ale (sample 10) <strong>and</strong><br />
the lagers (samples 13 <strong>and</strong> 14) exceeds the threshold. Acomparable threshold for maltose<br />
is not available; that quoted (1.36%) well exceeds the concentrations of maltose in the<br />
beers mentioned in Table 19.5 but, if the comparable threshold was similar to that of<br />
lactose (0.16%), as suggested in Table 20.3, the level of maltose in beers no. 1, 5, 6, 13,<br />
<strong>and</strong> 15 would exceed the threshold <strong>and</strong> influence the taste.<br />
Thetastethresholdforisohumulone(8.40)isreportedtobe5.6mg/l(1.5 10 5 M)<strong>and</strong><br />
that of hulupone (8.85) 7.7mg/l (2.3 10 5 M)(Gienapp <strong>and</strong> SchroÈder, 1975). Similarly,<br />
Weiss et al. (2002) found the best estimated threshold for the dicyclohexylamine salt of<br />
trans-iso- -acids (66.6% iso- -acids) in tap water was 4.54mg/l (1.25 10 5 M). Thus,<br />
with one exception, the level of iso- -acids (bitterness units) in all the beers analysed in<br />
Table19.1exceedsthetastethreshold,intheextremecaseby17times.However,thelevel<br />
ofhulupones reported(1.1±4.3mg/l) does not exceed thethreshold.By traditional scaling<br />
methods quinine hydrochloride was some six times more bitter than an iso- -acids<br />
preparation which, in turn, was about ten times more bitter than caffeine (Lewis et al.,<br />
1980). The character of the bitterness of these compounds is also different; iso- -acids<br />
especially, <strong>and</strong> quinine to aminor extent, were perceived as alingering bitterness on the<br />
back of the throat while caffeine gives ashort-lived bitterness on the tongue.<br />
Weiss et al. (2002) also found the best estimated threshold for tetrahydroiso- -acids<br />
(8.47)intapwaterwas1.61mg/l(4.3 10 6 M)butinunhoppedbeerwas8.50mg/l(2.32<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
10 5 M). Similarly, the best estimated threshold for the rho-iso- -acids (8.49) was 6.03mg/l<br />
(1.65 10 5 M) in tap water <strong>and</strong> 11.22mg/l (3.03 10 5 M) in commercial beer in<br />
agreement with Table 8.3. These values show the ability of beer to mask bitterness.<br />
The levels of glutamic acid (Tables 9.3 <strong>and</strong> 19.12) do not exceed the taste threshold<br />
given in Table 20.2 <strong>and</strong> the levels of 5 0 -inosine monophosphate <strong>and</strong> 5 0 -guanosine<br />
monophosphate (Table 19.13) in beer are below the taste thresholds but these compounds<br />
are reported to be flavour modifiers rather than primary flavours. Additions of 5 0 -<br />
guanosine monophosphate do modify beer flavour but the lowest level of addition of 5 0 -<br />
GMP required to alter beer flavour is greatly in excess of the amount naturally present<br />
(Clapperton, 1974).<br />
Despite their widespread adoption the use of thresholds to give flavour units has been<br />
criticized: `Thresholds are but one point on dynamic concentration continum' (Pangborn,<br />
1980). There is no evidence that intensity/concentration curves for all substances are<br />
parallel differing only in the point where they cross the abscissa. Further, taste is not a<br />
single instantaneous sensation but has atemporal element. Tasters have been trained to<br />
record the intensity on ascale between 0(none) <strong>and</strong> 100 (extreme) on amoving recorder<br />
chart whereby time-intensity curves such as Fig. 20.6 are obtained (Lewis et al., 1980).<br />
Normally the sample is held in the mouth for 10 seconds <strong>and</strong> then expectorated or, if<br />
beer, swallowed. As would be expected when asucrose gelatine was expectorated the<br />
intensity of the sweet sensation immediately started to fall <strong>and</strong> declined to zero in about<br />
10s. In contrast, when asample of beer was swallowed the intensity of the bitterness<br />
continued to rise for afurther 8sbefore starting to fall (Fig. 20.6).<br />
The bitter sensation appears to persist longer (60±90s) than the sweet sensation.<br />
Hughes <strong>and</strong> Bolshaw (1995) compared the time-intensity curves for trans-isohumulone,<br />
<strong>and</strong>preparationsofdihydro-,tetrahydro-,<strong>and</strong>hexahydro-iso- -acids(Fig.20.7).Thetwo<br />
tasters ranked the compounds in the same order; tetrahydroiso- -acids, hexahydroiso- -<br />
acids, trans-isohumulone, <strong>and</strong> dihydroiso- -acids, but the shape of the curves was<br />
different. Taster Bfound the tetrahydroiso- -acids much more bitter than the other<br />
compounds <strong>and</strong> the aftertaste persisted for over three minutes.<br />
Average intensity (n = 69)<br />
40<br />
30<br />
20<br />
10<br />
0<br />
30 ppm<br />
20<br />
10<br />
0<br />
0 10 20 30 40 50 60 70 80 90<br />
Time (s)<br />
Fig. 20.6 Average time-intensity curve for bitterness of four levels of iso- -acids in<br />
commercial lager. Samples were placed on the mouth at zero times <strong>and</strong> swallowed at 10 s. The<br />
judge continued to record intensity of bitterness until disappearance or for a maximum of two<br />
minutes. (after Lewis et al., 1980).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 20.7 Mean time-intensity curves of trans-isohumulone <strong>and</strong> mixtures of the chemically<br />
modified hop bitter acids. Samples were tested in 10 mM sodium phosphate buffer (pH4.15)<br />
with 0.05% (v/v) ethanol in black opaque glasses under red light. The solutions were 33.1 M<br />
with respect to the hop bitter acid (Hughes <strong>and</strong> Bolshaw, 1995).<br />
In general the olfactory threshold of acompound is several orders lower than the taste<br />
threshold (Table 20.4). This may represent the probability of acompound taken into the<br />
mouth reaching the olfactory epithelium. In atriangular test (see later) the assessors may<br />
be required to distinguish the similar beers first on the basis of odour. Indeed with one<br />
brewery taste panel it was found that most members differentiated between two beers on<br />
the basis of aroma rather than taste 90±95% of the time (Hoff et al., 1978). The olfactory<br />
threshold (OT) can be calculated from the simplified equation:<br />
log…OT† ˆ logK L=A 0:1Ao ‡…22:13 0:5†<br />
where KL/A is the absorption constant for molecules passing from air to the aqueous-lipid<br />
interface (usually between 6.0 <strong>and</strong> 8.5) <strong>and</strong> Ao is the cross-sectional area of the molecule<br />
(usually between 10±60 A Ê2 ).Thus the olfactory threshold of apure compound can be<br />
calculated to the first approximation from the partial pressure of the compound above an<br />
aqueous solution of the compound, its partition coefficient between water <strong>and</strong> light<br />
petroleum or octanol (substituting for the lipid membrane), <strong>and</strong> its cross-sectional area,<br />
which can be calculated from models.<br />
Foramoleculetoelicitaflavour itmustreach<strong>and</strong>reactwithaspecificreceptor<strong>and</strong>to<br />
do so it will probably have to pass through alipid membrane, thus the lipophilicity of the<br />
molecule will govern this approach. Gardner (1979) showed highly significant<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.4 Odour <strong>and</strong> taste thresholds of various compounds<br />
correlations between taste threshold <strong>and</strong> lipophilicity. Thus for ahomologous series of<br />
compounds (alcohols, esters, ketones, aldehydes <strong>and</strong> acids) in beers:<br />
log…1=T†ˆ log…P†‡b<br />
Odour threshold in water Taste threshold in beer<br />
(ppb) (Guadagni, 1970) (ppb) (Meilgaard, 1975)<br />
Ethanol 100,000 14,000,000<br />
Butyric acid 250 2,000<br />
Nootkatone 170 -<br />
Humulene 160 -<br />
Butanal 70 1 000<br />
Myrcene 15 ±<br />
n-Amyl acetate 5 5,000<br />
Dimethyl sulphide 0.3 50<br />
n-Decanal 0.1 6<br />
Methyl mercaptan 0.02 2.0<br />
-Ionone 0.007 1.3<br />
2-Methoxy-3-isobutylpyrazine 0.002 ±<br />
where Tˆthreshold in mol/l <strong>and</strong> Pˆoctanol/water partition coefficient representing<br />
lipophilicity. This relationship breaks down when Pis greater than 3.0. Similarly, in<br />
many series of compounds with bitter taste, the bitterness increases with increasing<br />
lipophilicity (Gardner, 1978). Accordingly, Kaneda et al. (2001, 2003) found that the<br />
absorption/desorption of beer on to alipid coated quartz crystal microbalance was related<br />
to the sensory bitterness. When amolecule has reached areceptor whether or not it<br />
initiatesasignaltothebraindependsuponitssize,shape,degreeofionization <strong>and</strong>charge<br />
pattern etc. The size <strong>and</strong> shape of molecules have also been expressed in terms of<br />
molecular connectivity (Kier <strong>and</strong> Hall, 1976).<br />
On the basis of threshold values <strong>and</strong> flavour units (FU), Meilgaard (1975) outlined the<br />
flavour chemistry of beer as illustrated in Table 20.5. Removal of any of the primary<br />
flavour constituents would produce adecisive change in flavour. Later work has not<br />
confirmed the importance of humuladienone in the hop aroma compounds but Goiris et<br />
al. (2002) have confirmed that an oxygenated sesquiterpenoid fraction is responsible for<br />
the spicy hop character of beer. Mackie <strong>and</strong> Slaughter (2002) have shown the importance<br />
of 2,5-dimethyl-4-hydroxy-3(2H)-furanone <strong>and</strong> related compounds among the caramelflavoured<br />
compounds. Removal of any one of the secondary constituents will produce a<br />
small change in flavour. Together the secondary flavour constituents form the bulk of a<br />
beer's flavour. Any differences between one beer <strong>and</strong> another of the same type is mostly<br />
determinedbyvariationsinthisclass.Tertiaryflavourconstituentsaddsubsidiaryflavour<br />
notes. Removal of any one of this class produces no perceptible change in flavour.<br />
Similarly it is not possible to say whether the numerous compounds, which individually<br />
contribute less than 0.1FU to the background flavour, are together of importance in beer<br />
flavour.<br />
Consideration of the concentration <strong>and</strong> threshold data in Table 19.8 will show which<br />
alcohols, acids <strong>and</strong> esters may contribute over 2FU in special beers <strong>and</strong> between 0.5 <strong>and</strong><br />
2.0FU in regular beers. Octanoic acid (difference threshold, 4.5mg/l), decanoic acid<br />
(1.5mg/l), dodecanoic acid (0.5mg/l), <strong>and</strong> to alesser extent hexanoic acid contribute to<br />
the caprylic (goaty) flavour in beer (Clapperton, 1978). The effect of the acids is additive<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.5 Tentative scheme for role of constituents in determining the flavour of beer<br />
(Meilgaard, 1975)<br />
1. Primary flavour constituents (above 2 FU*)<br />
Ethanol<br />
Hop bitter compounds (e.g., isohumulone)<br />
Carbon dioxide<br />
Speciality beers<br />
Hop aroma compounds (e.g., humuladienone)<br />
Caramel-flavoured compounds<br />
Several esters <strong>and</strong> alcohols (high gravity beers)<br />
Short-chain acids<br />
Defective beers<br />
2-trans-Nonenal (oxidized, stale)<br />
Diacetyl <strong>and</strong> 2,3-pentanedione (fermentation)<br />
Hydrogen sulphide, dimethyl sulphide <strong>and</strong> other compounds (fermentation)<br />
Acetic acid (fermentation)<br />
2-Methylbut-2-enylthiol (light struck-hops)<br />
Others (microbial infection etc.)<br />
2. Secondary flavour constituents (0.5±2.0 FU)<br />
Volatiles<br />
Banana esters (e.g., isoamyl acetate)<br />
Apple esters (e.g., ethyl hexanoate)<br />
Fusel alcohols (e.g., isoamyl alcohol)<br />
C6, C 8, C 10 aliphatic acids<br />
Ethyl acetate<br />
Butyric <strong>and</strong> isovaleric acids<br />
Phenylacetic acid<br />
Non-volatiles<br />
Polyphenols<br />
Various acids, sugars, hop compounds<br />
3. Tertiary flavour constituents (0.1±0.5 FU)<br />
2-Phenethyl acetate, o-aminoacetophenone<br />
Isovaleraldehyde, methional, acetoin<br />
4-Ethylguaiacol, gamma-valerolactone<br />
4. Background flavour constituents (below 0.1 FU)<br />
Remaining flavour compounds<br />
* Flavour Units (FU) = concentration/threshold<br />
<strong>and</strong> there is a linear relationship between the panel score for caprylic flavour <strong>and</strong><br />
concentration of octanoic + decanoic acids. During fermentation lager yeasts produce<br />
larger amounts of these acids than ale yeasts. Thus the caprylic flavour was observed in<br />
most of the lagers <strong>and</strong> 25% of the ales examined (Clapperton <strong>and</strong> Brown, 1978).<br />
20.3 Flavour stability<br />
Beer flavour is not static but in a continual state of change. The point where maturation<br />
ends <strong>and</strong> deterioration begins is undoubtedly different for different beers <strong>and</strong> probably<br />
different for different consumers. The off-flavour in one beer may be an essential<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Intensity<br />
Bitter taste<br />
Ribes aroma<br />
Sweet taste,<br />
toffee-like aroma<br />
<strong>and</strong> flavour<br />
Time<br />
Sweet aroma<br />
Cardboard flavour<br />
Fig. 20.8 Diagrammatic representation of sensory changes in beer flavour during ageing<br />
(Dalgliesh, 1977).<br />
character of another. Flavour stability has been reviewed by Dalgliesh (1977) <strong>and</strong> Tressl<br />
et al. (1980) have reviewed off-flavours. Since brewers can now largely control<br />
biological instability <strong>and</strong> non-biological haze, it is probably flavour instability that<br />
determines the shelf-life of the product. Some of the changes in flavour that occur during<br />
beer ageing are illustrated in Fig. 20.8.<br />
Many of the chemical changes during storage involve oxidation <strong>and</strong> will be<br />
accelerated if the beer is allowed to come in contact with oxygen after it leaves the<br />
fermentation vessel. The various electronic states of oxygen are illustrated in Fig. 20.9<br />
(Lacan et al., 2000). The ground state is adiradical with two unpaired electrons (having<br />
the same spin) in two * antibonding orbitals <strong>and</strong> is comparatively unreactive.<br />
Photoexcitation of the ground state gives `singlet oxygen' (with two electrons of opposite<br />
spin) but reduction gives asuperoxide radical ion (O2 ). This superoxide can be formed<br />
biochemically from oxygen by reactions involving ubiquinones, catecholamines or thiols<br />
<strong>and</strong> is normally converted to hydrogen peroxide (H2O2) by the enzyme superoxide<br />
dimutase (SOD). However, in acidic media it forms the hydroperoxy radical (HOO )<br />
which spontaneously disproportionates into hydrogen peroxide <strong>and</strong> oxygen. In the cell<br />
hydrogen peroxide is usually destroyed by the enzyme catalase but it can lead to even<br />
more `reactive oxygen species' (ROS), notably the hydroxy radical (HO ). Thus,<br />
hydrogen peroxide can react with superoxide (Haber-Weiss reaction) or with ferrous ions<br />
(Fenton reaction) (Fig. 20.10) The hydroxyl radical is probably responsible for initiating<br />
the autoxidation of lipids <strong>and</strong> the role of the Fenton reaction shows the importance of<br />
traces of heavy metals on beer deterioration. Kaneda et al. (1992) using electron spin<br />
Singlet<br />
1 O2<br />
hν<br />
Ground state<br />
O2<br />
Superoxide<br />
O 2– •<br />
radical anion<br />
Peroxide<br />
O 2–<br />
Fig. 20.9 Electronic configurations of various oxygen species (Lacan et al., 2000).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Haber-Weiss reaction<br />
O2 –<br />
+ H2O2 O2 + + HO –<br />
•<br />
HO•<br />
Superoxide Hydroxyl Hydroxyl<br />
radical ion<br />
Fenton’s reaction<br />
Fe 2+ H2O2 Fe 3+<br />
resonance (ESR) showed that the signal due to non-heme Fe 3‡ increased during beer<br />
storage. After the addition of potassium hydroxide another signal in the ESR spectrum<br />
indicated the presence of free radicals but the half-life of the hydroxy radical is reported<br />
as only 10 9 s.Uchida <strong>and</strong> Ono (1999) measured the increase in level of hydrogen<br />
peroxide in beer <strong>and</strong> found it paralleled the formation of hydroxy radicals.<br />
Unsaturated fatty acids, either free or esterified, are particularly susceptible to<br />
autoxidation. Compared with oleic acid (cis-octadec-9-enoic acid), linoleic acid (20.6,<br />
cis, cis-octadeca-9,12-dienoic acid) reacts with oxygen 12 times faster <strong>and</strong> linolenic acid<br />
(cis, cis, cis-octadeca-9,12,15-trienoic acid) reacts 25 times faster.<br />
13 12 11 10 9<br />
R1·CH CH·CH2·CH CH·R2<br />
R1 CH3(CH2)4·<br />
(20.6)<br />
(20.7)<br />
R1·CH•·CH CH·CH CH·R2 +O2/H• / R1·CH(OOH)·CH CH·CH CH·R2<br />
R1·CH CH·CH CH·CH•R2 +O2/H• / R1·CH CH·CH CH·CH(OOH)·R2<br />
·H•<br />
R1·CH CH·CH•·CH CH·R2<br />
R1·CH•·CH CH·CH CH·R2<br />
R1·CH CH·CH CH·CH•·R2<br />
+ + HO –<br />
HO•<br />
Fig. 20.10 Production of the hydroxyl radical.<br />
R2 ·(CH2)7COOR<br />
(20.8)<br />
(20.9)<br />
Autoxidation involves abstraction of ahydrogen radical from acarbon atom adjacent to a<br />
double bond, e.g., positions 8<strong>and</strong> 11 in oleic acid <strong>and</strong>, particularly, position 11, between<br />
two double bonds, in linoleic acid (20.6). This forms amesomeric radical (20.7) which<br />
then reacts with oxygen to form a mixture of hydroperoxy radicals (ROO ) which<br />
abstract another proton to form hydroperoxides (ROOH), e.g. (20.8) <strong>and</strong> (20.9). Thus,<br />
with linoleic acid, hydroperoxides at positions 9<strong>and</strong> 13 predominate. Double bonds that<br />
migrate usually adopt trans-geometry. Reduction of the hydroperoxide gives ahydroxyl<br />
group when hydroxylation of the double bonds gives isomers of trihydroxyoctadecenoic<br />
acids (Table 19.7). These acids are the precursor of trans-2-nonenal, which is responsible<br />
for the cardboard flavour in stale beer (Lermusieau et al., 1999; NoeÈl et al., 1999). It<br />
appears that this compound is not formed from oxygen in the head space of bottled beer<br />
but during wort preparation when it is bound, for example by proteins or sulphites, <strong>and</strong><br />
released during storage (NoeÈl et al., 1999).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Thethresholdoftrans-2-nonenalisabout0.1 g/l<strong>and</strong>ithasbeensuggestedthatitslevel<br />
provides aindication of the degree of staling. Other carbonyl compounds formed from the<br />
lipids in beer by irradiation with light include the C9, C10, <strong>and</strong> C11-alka-2,4-dienals<br />
(thresholds 0.5, 0.3 <strong>and</strong> 0.01ppb respectively) (Tressl et al., 1980). Using GC-O Evans et<br />
al. (1999) confirmed that during ageing the level of most aldehydes increased. As well as<br />
trans-2-nonenal, the levels of phenylacetaldehyde, methional, 4-methoxybenzaldehyde <strong>and</strong><br />
heptanal increased but the level of octanal fell. The level of diacetyl <strong>and</strong> pentane-2,3-dione<br />
in arange of commercial beers is given in Table 19.11. Quantities in excess of 0.15ppm<br />
impart abuttery flavour more noticeable in lagers than in ales. Bacterial contamination <strong>and</strong><br />
petite mutants of yeast result in high levels of these diketones.<br />
Although 18 Oin the headspace was not incorporated into the carbonyl fraction during<br />
storage of bottled beer, it was involved in the oxidation of sulphites, polyphenols <strong>and</strong><br />
isohumulones. Roughly 1ml of air in a300ml bottle will give an oxygen content of 1ppm,<br />
which is probably enough to oxidize all the reductones present in alight lager beer. The<br />
dissolved oxygen inbeer rapidly disappears, usually withoutthe immediate formationof an<br />
off-flavour, but the damage may have been done as beer contains compounds such as<br />
melanoidins <strong>and</strong> reductones, which act as oxygen carriers, <strong>and</strong> produce off-flavours at a<br />
later date. Chemically, reductones contain the grouping -C(OH)ˆC(OH)-CˆO- <strong>and</strong> the<br />
best characterized reductone is ascorbic acid (vitamin C)(4.96). Ascorbic acid has been<br />
detected in green malt <strong>and</strong> the leaves of green hops but is destroyed during kilning.<br />
Ascorbicacid,<strong>and</strong>otherreductones,readilycombinewithoxygen<strong>and</strong>accordinglyascorbic<br />
acid finds use as achill-proofing agent in beer (Chapter 15). Ascorbic acid (4.96) is<br />
reversibly oxidized to dehydroascorbic acid (4.97) but more extensive decomposition<br />
occurs under quite mild conditions. In model experiments designed to assess the efficiency<br />
of various substances to degrade valine to isobutyraldehyde by the Strecker mechanism,<br />
dehydroascorbic acid was 5 10 times more active than fructose which, in turn, was 2 3<br />
times more active than glucose or sucrose or ascorbic, pyruvic or chlorogenic acids (Swain<br />
<strong>and</strong> Casey, 1963). Ascorbic acid is usually estimated colorimetrically with the oxidationreduction<br />
indicator, 2,6-dichlorophenolindophenol (20.10), but other reductones will<br />
interfere. The Indicator Time Test (ITT) (Gray <strong>and</strong> Stone, 1939), which measures the<br />
decolorization of 2,6-dichlorophenolindophenol, gives an indication of the oxidationreduction<br />
or redox level of abeer.<br />
OH<br />
N<br />
Cl Cl<br />
O<br />
2,6-Dichlorophenolindophenol<br />
(20.10)<br />
The bitterness of beer declines during storage (Fig. 20.8). De Cooman et al. (2000)<br />
found that both in lagers <strong>and</strong> top-fermented beers the trans-iso- -acids deteriorated at a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Me3C<br />
OH<br />
Me<br />
(20.11) BHT<br />
CMe3<br />
+ R·OO•<br />
much faster rate than the cis-iso- -acids. During the first 15 months the loss of iso- -<br />
acids was mainly due to decomposition of the trans-isomers. In the trans-iso- -acids the<br />
double bonds in the 3-methyl-2-butenyl- <strong>and</strong> the 4-methyl-3-pentenoyl- side chains lie<br />
close together <strong>and</strong> it is suggested that this increases their susceptibility to autoxidation.<br />
Therefore the cis-trans ratio of the iso- -acids gives ameasure of abeer's deterioration.<br />
As discussed in Chapter 19 the flavours due to 4-vinylphenol <strong>and</strong> 4-vinylguaicol (4.134,<br />
threshold, 0.3 mg/l) are regarded as off-flavours in most beers but, with orcinol, are<br />
responsible for the clove-like character of Weizenbier.<br />
Beer also contains many antioxidants derived mainly from the polyphenols present in<br />
the malt <strong>and</strong> hops. Phenols react readily with free radicals but form mesomeric radicals<br />
which are not sufficiently energetic to propagate the free radical chain further (Lacan et<br />
al., 2000). For example, one molecule of the synthetic antioxidant, butylatedhydroxytoluene<br />
(BHT) (20.11) can destroy two hydroperoxy radicals (Fig. 20.11).<br />
Thelevelofsomevolatile sulphurcompoundsincreases during storage.We saw inthe<br />
last chapter that low levels of hydrogen sulphide are acceptable in ales <strong>and</strong> dimethyl<br />
sulphide is characteristic of some lagers (Table 19.20). As mentioned earlier, beers<br />
bottled in clear glass bottles <strong>and</strong> exposed to sunlight develop skunky `sunstruck' flavours<br />
due to 3-methyl-2-butenyl thiol (8.48, prenyl mercaptan) which has avery low threshold<br />
of 0.005ppb in water <strong>and</strong> 0.05ppb in beer. Other beers acquire aflavour described as<br />
`catty' or Ribes, as asimilar aroma is given off by the leaves <strong>and</strong> stems of flowering<br />
currants (Ribes spp.). The development of this flavour is closely correlated with the<br />
amount of headspace air (Clapperton, 1976). In beers bottled with high volumes of<br />
O<br />
SH<br />
Me3C<br />
8-Mercapto-p-menthan-3-one<br />
(20.12)<br />
O•<br />
O O<br />
Me<br />
Me OOR Me •<br />
CMe3<br />
Me3C CMe3 R·OO•<br />
Me3C CMe3<br />
Fig. 20.11 Antioxidant activity of Butylatedhydroxytoluene (BHT).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
headspace air the flavour develops rapidly over six weeks but thereafter slowly declines.<br />
One substance responsible for the catty flavour is 4-mercaptopentan-2-one (threshold,<br />
0.005ppb in water, 0.05ppb in beer). Beers with strong catty odours contained 1.5ppb of<br />
the mercaptopentanone but there may be other beer constituents which contribute to this<br />
off-flavour. Elsewhere (see Table 20.11 on page 750) 8-mercapto-p-menthan-3-one<br />
(20.12, p-menthane-8-thiol-3-one) has been proposed as a reference st<strong>and</strong>ard for the catty<br />
(Ribes) flavour.<br />
Gijs et al. (2002) applied Aroma Extract Dilution Analysis to fresh <strong>and</strong> aged beers<br />
(five days at 40 ëC, pH4.2) <strong>and</strong> found that the Flavour Dilution (FD) values were<br />
increased in the aged beer for ethyl butyrate, dimethyl trisulphide, 2-acetylpyrazine,<br />
methional, 2-methoxypyrazine, maltol (9.11), -nonalactone, -damascenone (8.165),<br />
<strong>and</strong> ethyl cinnamate. -Damascenone <strong>and</strong> an unknown compound, with a `dentist',<br />
smoked, vanilla odour with the same FD (243) are probably the most important odours in<br />
aged beer <strong>and</strong> more important than trans-2-nonenal (FD 81). Strecker degradation of the<br />
amino acid methionine gives methional, (3-(methylthio)propionaldehyde), which is<br />
responsible for the worty flavour of alcohol-free beers (PerpeÁte <strong>and</strong> Collin, 1999). It is<br />
also the precursor of dimethyl trisulphide which develops as beer ages (Gijs et al., 2000;<br />
Gijs <strong>and</strong> Collin, 2002).<br />
-Damascenone (8E-megastigma-3, 5, 8-trien-7-one, (8.165)), a degradation product<br />
of the carotenoid neoxanthin, is a key odour compound in a number of fruits <strong>and</strong> has an<br />
extremely low threshold (0.02 0.09 ng/g (ppb) in water). It is present in hops but is also<br />
found in unhopped wort. Hopped wort contained 450 ng/g but this was reduced during<br />
fermentation so fresh beers contained only low levels of -damascenone (6±25 ng/g).<br />
However, during ageing (five days at 40 ëC) the level increased to as much as 210 ng/g<br />
(Chevance et al., 2002). Experiments with -glucosidase suggest that the production of<br />
-damascenone during beer ageing can be partially explained by the hydrolysis of<br />
glucosides. This was confirmed as the production of -damascenone fell in beers aged at<br />
higher pH values (Gijs et al., 2002). The production of dimethyl trisulphide also fell at<br />
higher pH values but that of 3-(methylthio)propionaldehyde increased.<br />
20.4 Sensory analysis<br />
Sensory analysis uses the human senses to assess flavours. It has been discussed in books<br />
by Amerine et al. (1965), the International Organization for St<strong>and</strong>ardization (1983),<br />
Meilgaard et al. (1987), Piggott (1988) <strong>and</strong> Lawless <strong>and</strong> Heymann (1999). Analytica-<br />
EBC/ASBC give details for paired comparison tests, triangular tests, duo trio tests,<br />
determining the threshold of added substances, description analysis, ranking tests <strong>and</strong><br />
provide a flavour terminology. For other tests, for example, the `A' or `not A' test, the 2out-of-5<br />
test, the Scheffe paired comparison test <strong>and</strong> the multiple paired comparison test,<br />
reference should be made to the above texts.<br />
Analytica-EBC/ASBC first give a glossary of terms <strong>and</strong> definitions, then detailed<br />
instructions for carrying out sensory analysis. They suggest a layout for a medium-sized<br />
sensory evaluation area for brewery control work; consumer preference testing requires a<br />
much larger panel. The tasting room should be situated in a quiet area free from any<br />
smells. They suggest six booths 60±80 cm wide with a counter top 40±50 cm deep <strong>and</strong><br />
90 cm high. Dividers between the booths should project c. 46 cm <strong>and</strong> extend from the<br />
floor to the ceiling. Walls, floors <strong>and</strong> ceiling should be of smooth non-absorbing material<br />
of a pale neutral colour. The booths may be equipped with a hatch (sliding door, bread<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
ox or carousel) so the samples can be delivered from an adjacent preparation room. The<br />
booths should be odour free with a slight overpressure of air that has been filtered through<br />
charcoal. The air should be at 22 ëC <strong>and</strong> at 45±55% relative humidity. Shadow-free<br />
illumination at 70±80 foot c<strong>and</strong>les should be provided. Odourless water <strong>and</strong> salt-free<br />
crackers should be available to cleanse the palate. A small sink may be provided but, if<br />
so, the drains should be readily dismantled to allow for cleaning. The booth may be<br />
equipped with a computer station so that the panellists' observations can be analysed<br />
quickly <strong>and</strong> automatically without resource to the forms detailed below.<br />
Liquid samples are best provided in straight-sided cylindrical 250 ml (8 oz.) glasses<br />
which may be deeply coloured to mask differences in colour, clarity or foaming. The<br />
glasses should be washed in an odour-free detergent such as sodium hexametaphosphate<br />
not more than 12 hours before the test. 50±100 ml of sample should be served at the<br />
designated temperature. A sample temperature of 12 ëC is suitable for full perception of<br />
flavour, such as detecting faults or small differences. Mouth-feel <strong>and</strong> drinkability are best<br />
evaluated at a lower temperature, e.g., 8 ëC. For preference tasting, the beer should be<br />
tasted at the usual drinking temperature, making due allowance for warming which may<br />
occur between pouring <strong>and</strong> drinking. The number of samples presented to each assessor,<br />
their order <strong>and</strong> coding should be carefully monitored. The order of presentation should be<br />
balanced so that each sample appears in a given position an equal number of times. For<br />
example, the possible positions for three products A, B, <strong>and</strong> C to be compared in a<br />
ranking test are:<br />
ABC ACB BCA BAC CBA CAB<br />
so a multiple of six assessors should be chosen so that the six possible combinations can<br />
be presented an equal number of times.<br />
The order of presentation should be r<strong>and</strong>omized, e.g., by drawing sample cards from a<br />
bag or using a table of r<strong>and</strong>om numbers. Similarly, the samples should be coded with<br />
three-digit r<strong>and</strong>om numbers. Too many samples should not be presented at a single<br />
session to avoid sensory fatigue. Panel sessions should be held before meals preferably<br />
between 10 <strong>and</strong> 12 a.m. The assessors should be carefully instructed what is required of<br />
them. The amount of sample to be tasted, how long it is to be held in the mouth <strong>and</strong><br />
whether it should then be swallowed or expectorated should be clearly stated. The use of<br />
the scoresheet, any terminology <strong>and</strong> the interpretation of the scales used should be<br />
explained. Finally, the information sought in the test <strong>and</strong> the type of judgement/<br />
evaluation required, e.g., difference, descriptive, preference, acceptance should be stated.<br />
At the end of the test, at a location away from the tasting area, the codes may be disclosed<br />
<strong>and</strong> the assessors allowed to discuss the results among themselves <strong>and</strong>/or with the panel<br />
leader. It is the job of the panel leader to keep the panellists motivated <strong>and</strong> give them<br />
regular reports on their results.<br />
Sensory tests have two main applications: (i) those in which the primary aim is to<br />
describe the product <strong>and</strong> (ii) those in which the aim is to distinguish between two or more<br />
products. With regard to the latter it is important to distinguish between the need to know<br />
if there is a difference at all, the magnitude of that difference, the direction (or quality) of<br />
that difference, the effect of that difference, for example, with regard to preference, <strong>and</strong><br />
whether all or only part of the population detects a difference. Analytica-EBC/ASBC<br />
provide a simplified key to decide which tests are relevant to a given problem.<br />
For the selection <strong>and</strong> training of assessors, EBC/ASBC suggest that you interview <strong>and</strong><br />
screen two to three times the number of assessors required. The general health of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
c<strong>and</strong>idates should be checked, whether they are on regular medication, whether they<br />
smoke, <strong>and</strong> their ability to communicate verbally. Smokers are not automatically<br />
disqualified. C<strong>and</strong>idates can then be subjected to matching tests. For example, they are<br />
presented with four to six coded (A±F) but unidentified common flavours (sucrose,<br />
tartaric acid, caffeine, sodium chloride, tannic acid <strong>and</strong> ferrous sulphate) <strong>and</strong> asked to<br />
familiarize themselves with the tastes using as much odourless water as they like to<br />
cleanse their palates. They are then presented with the same samples labelled with threedigit<br />
numbers <strong>and</strong> asked to say which of the st<strong>and</strong>ards each resembles. C<strong>and</strong>idates<br />
scoring less than 80% matches should be rejected. For detection tests, c<strong>and</strong>idates are<br />
presented with three samples containing beer to which additions have been made versus<br />
control samples in a Triangular test (see later). The additions can be the above flavour<br />
st<strong>and</strong>ards, first at four to six times the threshold, <strong>and</strong> then at two to three times the<br />
threshold. C<strong>and</strong>idates scoring less than 60% at the higher level should be rejected <strong>and</strong><br />
preference given to those scoring 100% at the higher level <strong>and</strong> over 60% at the lower<br />
level. Sequential triangular tests can be used to find the taste acuity of the c<strong>and</strong>idates but<br />
a more important skill is the ability to detect (<strong>and</strong> grade) individual flavour notes in a<br />
`fog' of other impressions.<br />
For ranking tests, the c<strong>and</strong>idates are asked to discriminate a set of beer samples to<br />
which additions have been made <strong>and</strong> which are presented in a r<strong>and</strong>om order. For<br />
example; 0, 0.6, 1.2, <strong>and</strong> 1.8 mg/l geraniol. Other suggested materials are sucrose, sodium<br />
chloride, isoamyl acetate, dimethyl sulphide <strong>and</strong> acetic acid. Only c<strong>and</strong>idates that rank<br />
samples correctly, or invert only adjacent pairs, should be accepted. Finally, to test for<br />
descriptive ability, present sets of five to ten stimuli that are typical of the samples to be<br />
evaluated. Present the samples one at a time <strong>and</strong> ask the c<strong>and</strong>idate to describe his or her<br />
response. Suitable substance <strong>and</strong> concentrations are given in Analytica-EBC/ASBC.<br />
However, most brewing companies use their own employees as tasters <strong>and</strong> this leads to<br />
little or no scope for selecting people with good sensory ability. In addition, employees<br />
often miss panel sessions <strong>and</strong> cannot devote time for training because of other duties.<br />
Accordingly, one brewery company has recruited an external expert sensory panel. From<br />
70 applicants, ten were selected <strong>and</strong> employed for three three-hour sessions per week The<br />
members of the panel were trained as tasters, not just `beer' tasters, <strong>and</strong> were exposed to a<br />
wide range of foods <strong>and</strong> drinks. After training, the 95% confidence limits for this new<br />
panel were less than 5% whereas those for the old in-house panel were 20±30% (Hegarty<br />
et al., 2001).<br />
For training, Analytica-EBC/ASBC suggest that the assessors should be instructed to<br />
be objective <strong>and</strong> ignore likes <strong>and</strong> dislikes unless specifically asked for preference<br />
information. The assessors should normally proceed in the order appearance, odour, taste,<br />
<strong>and</strong> aftertaste. When assessing odour, the assessors should take short rather than long<br />
sniffs <strong>and</strong> not sniff too many times so they become fatigued <strong>and</strong> confused. As well as<br />
training in recognition <strong>and</strong> detection of basic tastes <strong>and</strong> odours, the panel should be<br />
trained in the use of descriptive language <strong>and</strong> scales. Samples should be presented to<br />
illustrate grainy, dry hop, kettle hop <strong>and</strong> overage flavours. The panel leader should<br />
analyse daily results <strong>and</strong> note for each assessor any obvious defects such as drift, lack of<br />
interest, or failure to detect obvious product variations. Such assessors should be called<br />
for special training. Most tasters will require attention at least three or four times a year.<br />
Because of the many opportunities for variability <strong>and</strong> bias resulting from the use of<br />
human subjects, reports of sensory tests should contain more detail than the reports of<br />
physical or chemical measurements. The report should begin with a summary, of not<br />
more than 110 words, answering the following:<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
· What was the objective?<br />
· What were the results?<br />
· What was done?<br />
· What can be concluded?<br />
Theobjectiveofthetestshouldbeclearlystatedwithacleardefinitionoftheproblem<strong>and</strong><br />
theapproachtakentosolveit.Thetestobjectivesshouldbeagreedbeforetheexperiment.<br />
Sufficient experimental detail should be given to allow the study to be repeated. The<br />
experimental design should be given showing how it meets the objectives, as should the<br />
sensory tests used, the measurements made, <strong>and</strong> the make up <strong>and</strong> previous experience of<br />
thepanel.Finally,detailsofsamplepreparation<strong>and</strong>presentation,theinformationgivento<br />
the panel <strong>and</strong> the statistical techniques used to analyse the results should be reported. The<br />
resultsshouldbepresentedintheformoftablesorfigures(notboth)<strong>and</strong>discussedbriefly<br />
stating whether they support or fail to support any original hypothesis. The report should<br />
end with clear-cut conclusions.<br />
The paired comparison test may be either (i) adirectional difference test or (ii) a<br />
paired preference test; it can also be used in assessor training. The former is used to<br />
determine in what way a particular sensory characteristic differs between the two<br />
samples, e.g., more sweet or less sweet. As arule at least seven assessors are required but<br />
up to 30 can be used (more for consumer tests). They should be familiar with the<br />
characteristic to be examined. For example, if the test concerns detection of an offflavour,<br />
the panel should first be allowed to taste asample free of any off-flavour <strong>and</strong>, if<br />
possible, a demonstration of the off-flavour. In general the inclusion of controls<br />
(reference substances) is advisable. Paired samples are offered simultaneously, an equal<br />
number of AB <strong>and</strong> BA with r<strong>and</strong>om three-digit codes. The assessors should be instructed<br />
to examine sets in aspecified order, e.g., always from left to right, however, they may<br />
make repeated tests of any sample while tasting is in progress. Specimen answer forms<br />
are given in Fig. 20.12. The test supervisor may use one of the following two<br />
possibilities:<br />
1. adopt the `forced choice' technique in which the assessor is asked to choose one<br />
sample or the other (by guessing if no difference is perceived) or<br />
2. allow the answer `no difference'.<br />
In most test situations (two-sided) the test question does not distinguish between the<br />
samples <strong>and</strong> the reply may favour one or the other sample. One-sided tests are<br />
occasionally used when the characteristic can vary only in one direction. The results are<br />
collected <strong>and</strong> interpreted by reference to Fig. 20.13 <strong>and</strong> Tables 20.6 <strong>and</strong> 20.7. With the<br />
`forced choice' technique asignificant difference or preference is established if the<br />
number found is equal to or larger than that given in the table. The `no difference'<br />
technique is not amenable to formal statistical analysis but two approaches can be used.<br />
In one the `no difference' replies are ignored <strong>and</strong> in the other half the `no difference'<br />
replies are allocated to each category. The test report should allow full identification of<br />
the samples examined, the characteristic studied, whether or not reference substances<br />
were used, <strong>and</strong> the results with their probability levels.<br />
The triangular test is used to determine whether asensory difference is apparent<br />
between two samples. The assessor is presented with aset of three samples, two of which<br />
are identical. After tasting the assessors complete aform similar to Fig. 20.14 showing<br />
the sample perceived to be different <strong>and</strong> the results are interpreted by reference to Table<br />
20.8. `No difference' results should be considered invalid. In the duo trio test the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Directional Difference Test<br />
Name ______________________ Date __________________<br />
day/month/year<br />
Object of Test _______________________________________<br />
Test Criterion _______________________________________<br />
Test Pairs Which sample is more __________________?<br />
Sample No. Sample No.<br />
_________ _________ __________________<br />
_________ _________ __________________<br />
_________ _________ __________________<br />
Comments<br />
Paired Preference Test<br />
Name ______________________ Date __________________<br />
day/month/year<br />
Object of Test _______________________________________<br />
Test Criterion _______________________________________<br />
Test Pairs Which sample is more __________________?<br />
Sample No. Sample No.<br />
_________ _________ __________________<br />
_________ _________ __________________<br />
_________ _________ __________________<br />
Comments<br />
Fig. 20.12 Specimen answer forms for the directional difference test <strong>and</strong> the paired preference<br />
test (Analytica-EBC).<br />
Directional difference test<br />
Paired preference test<br />
Two-sided test<br />
Question: which sample has<br />
the stronger intensity of the<br />
charcteristics studied?<br />
Count the number of replies<br />
citing one of the two samples<br />
the more frequently.<br />
Conclude that the intensity for<br />
this sample is significantly<br />
stronger than for the other if<br />
the number obtained is greater<br />
than or equal to that shown in<br />
Table 20.7.<br />
Question: Which sample do<br />
you prefer?<br />
Count the number of replies<br />
citing one of the two samples<br />
the more frequently.<br />
Conclude that this sample is<br />
significantly preferred to the<br />
other if the number obtained<br />
is greater than or equal to that<br />
shown in Table 20.7.<br />
One-sided test<br />
Question: which sample has<br />
the stronger intensity of the<br />
charcteristics studied?<br />
Count the number of replies<br />
choosing the sample of interest.<br />
Conclude that this stronger<br />
intensity is significantly<br />
apparent if the number of<br />
positive replies is greater than<br />
or equal to the number shown<br />
in Table 20.6.<br />
Question: Which sample do<br />
you prefer?<br />
Count the number of replies<br />
choosing the sample of interest.<br />
Conclude that there is a preference<br />
for the sample of interest<br />
if the number of positive replies<br />
is greater than or equal to the<br />
number shown in Table 20.6.<br />
Fig. 20.13 Questions <strong>and</strong> interpretations of the directional difference test <strong>and</strong> the paired<br />
preference test (Analytica-EBC).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.6 One-sided test (P =0.50 with nreplies)<br />
Number of Minimum number of positive replies for significance level of<br />
replies<br />
0.05 0.01 0.001<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
14<br />
15<br />
16<br />
17<br />
18<br />
19<br />
20<br />
21<br />
22<br />
23<br />
24<br />
25<br />
30<br />
35<br />
40<br />
45<br />
50<br />
60<br />
70<br />
80<br />
90<br />
100<br />
7 7 ±<br />
7 8 ±<br />
8 9 ±<br />
9 10 10<br />
9 10 11<br />
10 11 12<br />
10 12 13<br />
11 12 13<br />
12 13 14<br />
12 14 15<br />
13 14 16<br />
13 15 16<br />
14 15 17<br />
15 16 18<br />
15 17 18<br />
16 17 19<br />
16 18 20<br />
17 19 20<br />
18 19 21<br />
20 22 24<br />
23 25 27<br />
26 28 31<br />
29 31 34<br />
32 34 37<br />
37 40 43<br />
43 46 49<br />
48 51 55<br />
54 57 61<br />
59 63 66<br />
The values given in the tables were calculated from the exact formula binomial law for parameter P=0.50 with<br />
n repetitions (replies).<br />
When the number of replies is higher than 100 use the following formula based on the approximation of the<br />
binomial law by the normal law which gives the actual minimum number of assessments to be obtained with a<br />
maximum error equal to at most one unit. Minimum number of replies: nearest whole value to (n + 1)/2 + k p n<br />
in which k is chosen from the table below.<br />
Level of significance K<br />
One-sided Two-sided<br />
0.05 0.82 0.98<br />
0.01 1.16 1.29<br />
0.001 1.55 1.65<br />
Tables for calculating other significance levels may be found in Roessler et al. (1978) <strong>and</strong> Fern<strong>and</strong>us et al.<br />
assessors are first presented with the identified reference sample. This is followed by two<br />
coded samples, one of which isidentical to the reference sample. The assessor is asked to<br />
identify the odd sample <strong>and</strong> complete the form Fig. 20.15 <strong>and</strong> the results are interpreted<br />
by reference to Table 20.9.<br />
The measurement of the threshold of added substances is not carried out routinely but<br />
is required when anew substance is found which may or may not influence the flavour of<br />
beer. The experimental design used is known in psychophysics as the forced choice<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.7 Two-sided test* (P =0.50 with nreplies)<br />
Number of replies Minimum number of positive replies for significance level of<br />
0.05 0.01 0.001<br />
7 7 ± ±<br />
8 8 8 ±<br />
9 8 9 ±<br />
10 9 10 ±<br />
11 10 11 11<br />
12 10 11 12<br />
13 11 12 13<br />
14 12 13 14<br />
15 12 13 14<br />
16 13 14 15<br />
17 13 15 16<br />
18 14 15 17<br />
19 15 16 17<br />
20 15 17 18<br />
21 16 18 19<br />
22 17 18 19<br />
23 17 19 20<br />
24 18 19 21<br />
25 18 20 21<br />
30 21 23 25<br />
35 24 26 28<br />
40 27 29 31<br />
45 30 32 34<br />
50 33 35 37<br />
60 39 41 44<br />
70 44 47 50<br />
80 50 52 56<br />
90 55 58 61<br />
100 61 64 67<br />
*Refer to footnotes of Table 20.6 (from Analytica-EBC (1998)).<br />
Name ____________________ Date ___________________<br />
day/month/year<br />
Product submitted to test ______________________________<br />
Problem: 3 samples are presented to you; circle the number of<br />
that which is different from the other 2.<br />
Comments<br />
Set of 3 samples<br />
Fig. 20.14 Specimen answer form for the triangular test (Analytica-EBC).<br />
modification of the ascending method of limits test. Sixteen or more assessors receive six<br />
sets of three beers each consisting of two controls <strong>and</strong> one test sample. Test samples<br />
increase in concentration by aconstant factor, usually 2.0. An approximate threshold (t a)<br />
may be determined first using five to ten assessors <strong>and</strong> increasing the concentration by a<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.8 Minimum number of correct replies to establish significance at various probability<br />
levels for the triangular test (one-sided p = 1/3)*<br />
Number<br />
of replies<br />
5<br />
6<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
14<br />
15<br />
16<br />
17<br />
18<br />
19<br />
20<br />
21<br />
22<br />
23<br />
24<br />
25<br />
26<br />
27<br />
28<br />
29<br />
30<br />
31<br />
32<br />
33<br />
34<br />
35<br />
36<br />
37<br />
38<br />
39<br />
40<br />
41<br />
42<br />
43<br />
44<br />
45<br />
46<br />
47<br />
48<br />
49<br />
50<br />
51<br />
52<br />
Minimum number of correct replies<br />
for a significance level of<br />
Number<br />
of replies<br />
Minimum number of correct replies for<br />
a significance level of<br />
0.05 0.01 0.001 0.05 0.01 0.001<br />
4 5 ±<br />
5 6 ±<br />
5 6 7<br />
6 7 8<br />
6 7 8<br />
7 8 9<br />
7 8 10<br />
8 9 10<br />
8 9 11<br />
9 10 11<br />
9 10 12<br />
9 11 12<br />
10 11 13<br />
10 12 13<br />
11 12 14<br />
11 13 14<br />
12 13 15<br />
12 14 15<br />
12 14 16<br />
13 15 16<br />
13 15 17<br />
14 15 17<br />
14 16 18<br />
15 16 18<br />
15 17 19<br />
15 17 19<br />
16 18 20<br />
16 18 20<br />
17 18 21<br />
17 19 21<br />
17 19 22<br />
18 20 22<br />
18 20 22<br />
19 21 23<br />
19 21 23<br />
19 21 24<br />
20 22 24<br />
20 22 25<br />
20 23 25<br />
21 23 26<br />
21 24 26<br />
22 24 27<br />
22 24 27<br />
22 25 27<br />
23 25 28<br />
23 26 28<br />
24 26 29<br />
24 27 30<br />
53<br />
54<br />
55<br />
56<br />
57<br />
58<br />
59<br />
60<br />
61<br />
62<br />
63<br />
64<br />
65<br />
66<br />
67<br />
68<br />
69<br />
70<br />
71<br />
72<br />
73<br />
74<br />
75<br />
76<br />
77<br />
78<br />
79<br />
80<br />
81<br />
82<br />
83<br />
84<br />
85<br />
86<br />
87<br />
88<br />
89<br />
90<br />
91<br />
92<br />
93<br />
94<br />
95<br />
96<br />
97<br />
98<br />
99<br />
100<br />
24 27 30<br />
25 27 30<br />
25 28 30<br />
26 28 31<br />
26 28 31<br />
26 29 32<br />
27 29 32<br />
27 30 33<br />
27 30 33<br />
28 30 33<br />
28 31 34<br />
29 31 34<br />
29 32 35<br />
29 32 35<br />
30 33 36<br />
30 33 36<br />
31 33 36<br />
31 34 37<br />
31 34 37<br />
32 34 38<br />
32 35 38<br />
32 35 39<br />
33 36 39<br />
33 36 39<br />
34 36 40<br />
34 37 40<br />
34 37 41<br />
35 38 41<br />
35 38 41<br />
35 38 42<br />
36 39 42<br />
36 39 43<br />
37 40 43<br />
37 40 44<br />
37 40 44<br />
38 41 44<br />
38 41 45<br />
38 42 45<br />
39 42 46<br />
39 42 46<br />
40 43 46<br />
40 43 47<br />
40 44 47<br />
41 44 48<br />
41 44 48<br />
41 45 48<br />
42 45 49<br />
42 46 49<br />
The values in this table were calculated from the exact formula: binomial law for parameter p = 1/3 with n<br />
repetitions (replies). When the number of replies is larger than 100, numbers of required correct replies<br />
may be obtained from the following formula based on the approximation binomial law by the normal law,<br />
with a maximum error equal to one unit: x ˆ 0:4714z n<br />
p ‡ ‰…2n ‡ 3†=6Š where z ˆ 1:64 for 0:05, 2.3<br />
for 0:01, <strong>and</strong> 3.10 for 0:001. The minimum number of correct replies is x if x is a whole number<br />
or the next higher integer if x is not a whole number. Tables for significance levels other than those listed<br />
here may be found in Amer. Soc. Testing <strong>and</strong> Materials (1979) <strong>and</strong> Jones (1956) (from Analytica-EBC<br />
(1988)).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.9 Significance of results in `one-sided test' (P ˆ 0.50 with n replies)<br />
Number of replies Minimum number of positive replies for significance level of<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
14<br />
15<br />
16<br />
17<br />
18<br />
19<br />
20<br />
21<br />
22<br />
23<br />
24<br />
25<br />
30<br />
35<br />
40<br />
45<br />
50<br />
60<br />
70<br />
80<br />
90<br />
100<br />
0.05 0.01 0.001<br />
7 7 ±<br />
7 8 ±<br />
8 9 ±<br />
9 10 10<br />
9 10 11<br />
10 11 12<br />
10 12 13<br />
11 12 13<br />
12 13 14<br />
12 14 15<br />
13 14 16<br />
13 15 16<br />
14 15 17<br />
15 16 18<br />
15 17 18<br />
16 17 19<br />
16 18 20<br />
17 19 20<br />
18 19 21<br />
20 22 24<br />
23 25 27<br />
26 28 31<br />
29 31 34<br />
32 34 37<br />
37 40 43<br />
43 46 49<br />
48 51 55<br />
54 57 61<br />
59 63 66<br />
The values given in the tables were calculated from the exact formula binomial law for the parameter P = 0.50<br />
with n repetitions (replies). When the number of replies is larger than 100, numbers of correct replies may be<br />
obtained from the following formula based on the approximation of the binomial law by the normal law, with a<br />
maximum error equal to one unit: x = (n + 1)/2 + k n<br />
p where k = 0.82 for 0.05: <strong>and</strong> k = 1.16 for 0.01;<br />
<strong>and</strong> 1.55 for 0.001. Tables for significance levels other than those listed here may be found in Roessler et al.<br />
(1978) <strong>and</strong> Fern<strong>and</strong>us et al. (1970) (from EBC-Analytica (1998)).<br />
Name ____________________ Date ___________________<br />
day/month/year<br />
Product submitted to test ______________________________<br />
Problem: The sample on the left is a control. Of the other 2<br />
samples, one is the same as the control <strong>and</strong> the other is different.<br />
Indicate the different sample.<br />
Comments<br />
Set of 3 samples<br />
Fig. 20.15 Specimen answer form for the triangular test (duo trio) (Analytica-EBC).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
0.25ta<br />
0.5ta<br />
Step no: 1 2 3 4 5 6<br />
ta<br />
Fig. 20.16 Threshold of added substances: example of the presentation of the six triangles t a=<br />
approximate threshold (Analytica-EBC).<br />
factor of 3.0. The dilution series in the main test is then: step 1, 0.25 ta; step 2, 0.5 ta;step<br />
3, ta; step 4, 2ta; step 5, 4ta; <strong>and</strong> step 6, 8ta set out as illustrated in Fig. 20.16. The<br />
assessors are asked to indicate the position of the test sample in each set of three beers.<br />
An example of the questionnaire is given in Fig. 20.17. For each assessor the best<br />
estimated threshold (BET) is calculated as the geometric mean of the highest<br />
concentration missed <strong>and</strong> the next highest adjacent concentration. Ahistogram of the<br />
individualBETsofthegroup isproduced<strong>and</strong>fromitthegroupthresholdisthegeometric<br />
mean of the BETs. In the extended form of the test, critical sets of three beers are<br />
repeated two to four times until both the assessor <strong>and</strong> the test supervisor agree that the<br />
threshold has been successfully bracketed by the assessor. For assessors at the top <strong>and</strong><br />
bottom of the range, extra concentration steps may be required.<br />
For description analysis 15±30 trained assessors are required. In apreliminary step,<br />
assessors agree which attributes (usually 10 to 40) will be used, <strong>and</strong> ascale is defined for<br />
each attribute, if possible with the use of reference st<strong>and</strong>ards. In the test itself, after<br />
tasting the sample, assessors award an intensity score for each attribute. Results may be<br />
used to form asensory profile of the sample, <strong>and</strong> profiles of two or more samples may be<br />
compared using statistical techniques. The attributes to be rated may be chosen from the<br />
122terms<strong>and</strong> the14 flavour classesofthe st<strong>and</strong>ardterminology(Table20.11). Tento20<br />
attributes are chosen for the simple descriptive test <strong>and</strong> up to 50 or more for the full test.<br />
2ta<br />
Ascending method of limits<br />
4ta<br />
Date ____________________ Assessor __________________<br />
8ta<br />
You have received 6 sets of beer samples. Each set is a triangle<br />
consisting of 2 identical controls <strong>and</strong> 1 test sample containing an<br />
added substance. Concentrations of the added substances increase<br />
from left to right. Please locate as many as you can on the test<br />
samples, indicating their position with a check mark in the<br />
corresponding box in each column. Avoid sensory fatigue: locate<br />
strong samples by smell or by taking very small sips, conserving<br />
your discriminatory power for those sets of 3 beers near your<br />
threshold. Review your results with a test supervisor.<br />
Triangle no.: 1 2 3 4 5 6<br />
Describe the flavour of the added substance _________________<br />
____________________________________________________<br />
Fig. 20.17 Threshold of added substances; example of questionnaire (Analytica-EBC).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Examples of intensity scales of proven usefulness are:<br />
Scale A<br />
or<br />
Scale C<br />
0 1 2 3 4 5 6 8 7 9<br />
0<br />
not<br />
present<br />
1<br />
2<br />
just slight<br />
recognizable<br />
3<br />
moderate<br />
A 15 cm line with decriptive terms 1.5 cm from each end<br />
4<br />
strong<br />
5<br />
very<br />
strong<br />
Scale B<br />
weak 0 0 0 0 0 0 0 strong<br />
weak strong<br />
Assessors place a mark on the line to indicate intensity. A numerical score can be<br />
obtained by measuring the distance, in cm, from the left h<strong>and</strong> end of the line to the mark.<br />
Reference st<strong>and</strong>ards should be used where possible to anchor two or three points on the<br />
scale.<br />
Panel scores<br />
Panel scores<br />
Panel scores<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
(i) Lager beer<br />
Aroma Flavour After<br />
palate<br />
2 3 6 12 15 17 40 1 2 3 12 15 17 23 25 26 29 30 33 34 39 40 2 39<br />
(ii) Pale ale<br />
2 3 6 12 1 2 3 6 12<br />
(iii) Draught bitter<br />
2 6 12 17 40 1 2 3 4 12 23<br />
23 25 26 29 30 31 33 34 39 2 39<br />
29 31 33 34 39 40 2 4 39 40<br />
Fig. 20.18 Description analysis; profile analysis of the effects on flavour of the addition of 0.2, 0.6<br />
<strong>and</strong> 1.8 mg/l of diacetyl to three different beers. Histograms show the scoring of intensity of aroma<br />
<strong>and</strong> flavour qualities by a panel of eight assessors. Scores are entered from left to right for the<br />
control beer <strong>and</strong> the corresponding beers resulting from the three increasing levels of addition of<br />
diacetyl. Continuous lines depict the profile descriptions of the flavours of the unadulterated beers.<br />
Key to qualities: 1 liveliness (CO 2 tingle), 2 sweet, 3 sickly, 4 toffee-like, 6 estery, 12 diacetyl,<br />
15 cabbagy, vegetable water, 17 sulphury, 23 smooth, 25 acidic (sharp), 26 sour, 29 mouth<br />
coating, 30 astringent, 31 drying, 33 body, 34 watery, 39 bitter, <strong>and</strong> 40 hoppy (Analytica-EBC).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Body<br />
Br<strong>and</strong> 1<br />
Br<strong>and</strong> 2<br />
Aroma<br />
strength<br />
Aftertaste Fruity, estery<br />
Bitterness<br />
Diacetyl<br />
Sulphidic/<br />
Tic<br />
Hop<br />
character<br />
Malty/grainy<br />
Fig. 20.19 Description analysis: example of spider web plot (Analytica-EBC). Spider web plot<br />
showing two different beers. No. of assessors, n = 26. Mean scores are shown as distance from<br />
the centre. The width of the line is the last significant difference about the mean calculated as<br />
the studentized range, SR. For example, the following ANOVA table was produced for the hop<br />
character means of two br<strong>and</strong>s.<br />
Source df SS MSE F-ratio<br />
Total 51 664.32 ± ±<br />
Between br<strong>and</strong>s 1 64.32 64.32 5.63<br />
Error 50 600.00 12.00 ±<br />
The error b<strong>and</strong>s would be calculated by finding<br />
SR ˆ Q MSE<br />
r<br />
ˆ 2:85<br />
N<br />
12<br />
r<br />
ˆ 1:933<br />
26<br />
Where Q is the upper 5 percentage points for two treatments <strong>and</strong> 50 degrees of freedom (from<br />
Malek et al., 1982).<br />
For each attribute the average rating is calculated <strong>and</strong> the results presented either as a<br />
Table or as ahistogram, for example as in Fig. 20.18. An alternative is aspider web plot<br />
(Fig. 20.19) where mean scores are shown as the distance from the centre. The ranking<br />
test is used to place aseries of test samples (usually from 3to 6) in rank order according<br />
to agiven characteristic (criterion). The criterion may be the intensity of asingle sensory<br />
attribute, or agroup of related attributes, or atotal impression. The test is especially<br />
suitable in those situations where scale estimates are not meaningful <strong>and</strong> it is convenient<br />
to rank aseries of samples according to preference or some other criteria. Assessors (n)<br />
receive the test samples (k) simultaneously in r<strong>and</strong>om order <strong>and</strong> rank them according to<br />
the specified criterion. The rank sums (R) are calculated <strong>and</strong> evaluated statistically with<br />
the aid of Friedman's test. Specimen answer forms are given in Figs 20.20 <strong>and</strong> 20.21. In<br />
preference tests, assessors are instructed to assign rank 1to the preferred sample, rank 2<br />
to the next preferred, etc. For intensity tests, assessors are instructed to assign rank 1to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the lowest intensity, rank 2 to the next lowest etc. Friedman's F can be calculated by the<br />
formula:<br />
F ˆ ‰…R1 R† 2 ‡ . . . …Rk R† 2 Š<br />
kR=6 A<br />
where R is the mean rank sum calculated as<br />
R ˆ …R1 . . . ‡ Rk†=k ˆ n…k ‡ 1†<br />
Name ___________________________ Date _____________<br />
(day/month/year)<br />
Products submitted to test _____________________________<br />
Problem: You have received 5 samples placed as follows:<br />
left mid left mid mid right right<br />
Taste the samples from left to right <strong>and</strong> write ‘1’ in the box of the<br />
sample you like best, ‘2’ in the box of the sample you like next best,<br />
<strong>and</strong> so on.<br />
Comments:<br />
Fig. 20.20 Specimen answer for a ranking test (preference) (Analytica-EBC).<br />
Ranking Test<br />
Name ______________________________ Date ______________<br />
(day/month/year)<br />
Products submitted to test __________________________________<br />
Problem: You have received 4 samples labelled with the three-digit<br />
numbers shown in the column marked ‘Samples presented’. Taste the<br />
samples <strong>and</strong> place them in rank order according to bitterness, listing<br />
the most bitter in the column marked 4, the next most bitter in the<br />
column marked 3, etc. If two samples appear the same, preferably make<br />
a ‘best guess’ as to their rank order, or if you cannot guess, indicate<br />
under “Comments” the sample numbers that could not be differentiated.<br />
Samples presented Order of rank Comments<br />
1. 2. 3. 4.<br />
set *1<br />
149, 251, 347, 428, 347, 428, 251, 149 428 = ~ 251<br />
set *2<br />
014, 017, 146, 155, 017, 146, 155, 014<br />
set *3<br />
098, 123, 233, 473, 233, 123, 473, 098 123 = ~ 473 = ~ 098<br />
Comments:<br />
4 products were compared in set *1, then presented again in sets 2 <strong>and</strong><br />
3 with different codes. Text items which are underlined were filled in<br />
by the panel leader before the test. Text in italics is the assessor’s response.<br />
The codes used for this assessor were: Sample A = 347, 146, 233;<br />
Sample B = 251, 017, 473; Sample C = 428, 014, 098; Sample D = 149,<br />
155, 123.<br />
Fig. 20.21 Specimen answer for a ranking test (bitterness) (Analytica-EBC).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.10 Ranking test; upper probability points for the<br />
No. of samples (k) No. of degrees of freedom of<br />
(df = n 1)<br />
3<br />
4<br />
5<br />
6<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
14<br />
15<br />
16<br />
17<br />
18<br />
19<br />
20<br />
21<br />
22<br />
23<br />
24<br />
25<br />
26<br />
27<br />
28<br />
29<br />
30<br />
31<br />
From EBC-Analytica (1998).<br />
2 with<br />
when nis the number of assessments. If Fexceeds the upper critical value of<br />
k 1degrees of freedom (Table 20.10) it can be concluded that there is asignificant<br />
difference between the samples. If practical it is best to prohibit ties as the statistics<br />
become cumbersome. In the equation above Ais an adjustment for ties; if no ties are<br />
present A ˆ 0. If ties are present consult Analytica-EBC/ASBC for the statistical<br />
treatment. The multiple comparison procedure according to Friedman is also given. The<br />
Least Significant Difference (LSD) within the set of rank sums is given by the formula:<br />
…<br />
p kR=3†<br />
LSDrank ˆt =2:00<br />
2 -distribution<br />
where t /2.00 is Student's t, which equals 1.96 at the 5% level <strong>and</strong> 2.58 at the 1% level of<br />
significance. Any two rank sums which differ by more than the LSD are significantly<br />
different.<br />
Finally, Analytica-EBC/ASBC describe an internationally accepted flavour terminologyfor<br />
beer.It names<strong>and</strong> defines each of122 separately identifiableflavour notes which<br />
can occur in beer (Table 20.11) The terminology was based on the principles that:<br />
1. Each separately identifiable flavour characteristic has its own name.<br />
2. Similar flavours are placed together.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
2<br />
3<br />
4<br />
5<br />
6<br />
7<br />
8<br />
9<br />
10<br />
11<br />
12<br />
13<br />
14<br />
15<br />
16<br />
17<br />
18<br />
19<br />
20<br />
21<br />
22<br />
23<br />
24<br />
25<br />
26<br />
27<br />
28<br />
29<br />
30<br />
2<br />
level of significance, %<br />
0.05 0.01<br />
5.99 9.21<br />
7.81 11.34<br />
9.49 13.28<br />
11.07 15.09<br />
12.59 16.81<br />
14.07 18.47<br />
15.51 20.09<br />
16.92 21.67<br />
18.31 23.21<br />
19.67 24.72<br />
21.03 26.22<br />
22.36 27.69<br />
23.68 29.14<br />
25.00 30.58<br />
26.30 32.00<br />
27.59 33.41<br />
28.87 34.80<br />
30.14 36.19<br />
31.41 37.57<br />
32.67 38.93<br />
33.92 40.29<br />
35.17 41.64<br />
36.41 42.98<br />
37.65 44.31<br />
38.88 45.64<br />
40.11 46.96<br />
41.34 48.28<br />
42.56 49.59<br />
43.77 50.89
Table 20.11 Description of the terminology system<br />
Particular relevance: O ˆ Odour T ˆ Taste M ˆ Mouth-feel W ˆ Warming Af ˆ Afterflavour<br />
Class term First tier Second tier Relevance Comments, synonyms, definitions Reference st<strong>and</strong>ard<br />
Class 1: aromatic, fragrant, fruity, floral<br />
0110 Alcoholic OTW General effect of ethanol <strong>and</strong> higher alcohols Ethanol, 50 g/1<br />
0111 Spicy OTW Allspice, nutmeg, peppery, eugenol: see also 1003 Vanilla Eugenol, 120 g/1<br />
0112 Vinous OTW Bouquet, fusely, wine-like (white wine)<br />
0120 Solvent-like OT Like chemical solvents<br />
0121 Plastics OT Plasticizers<br />
0122 Can-liner OT Lacquer-like<br />
0123 Acetone OT (Acetone)<br />
0130 Estery OT Like aliphatic esters<br />
0131 Isoamyl acetate OT Banana, pear drop (Isoamyl acetate)<br />
0132 Ethyl hexanoate OT Apple-like with note of aniseed: see also 0142 Apple (Ethyl hexanoate)<br />
0133 Ethyl acetate OT Light fruity solvent-like: see also 0120 Solvent-like (Ethyl acetate)<br />
0140 Fruity OT Of specific fruits or mixtures of fruits<br />
0141 Citrus OT Citral, grapefruit, lemony, orange rind<br />
0142 Apple OT<br />
0143 Banana OT<br />
0144 Blackcurrant OT Blackcurrant fruit; for blackcurrant leaves use 0810 Catty<br />
0145 Melony OT (6-Nonenal, cis-or trans-)<br />
0146 Pear OT<br />
0147 Raspberry OT<br />
0148 Strawberry OT<br />
0150 Acetaldehyde OT Green apples, raw apple skin, bruised apples (Acetaldehyde)<br />
0160 Floral OT Like flowers, fragrant<br />
0161 2-Phenylethanol OT Rose-like (2-Phenylethanol)<br />
0162 Geraniol OT Rose-like, different from 0161; taster should compare pure chemicals (Geraniol)<br />
0163 Perfumy OT Scented (Exaltolide musk)<br />
0170 Hoppy OT Fresh hop aroma; use with other terms to describe stale hop aroma;<br />
does not include hop bitterness (see also 1200 Bitter)<br />
0171 Kettle hop OT Flavour imparted by aroma hops boiled in kettle<br />
0172 Dry-hop OT Flavour imparted by dry hops added in tank or cask<br />
0173 Hop oil OT Favour imparted by addition of distilled hop oil<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.11 Continued<br />
Class term First tier Second tier Relevance Comments, synonyms, definitions Reference st<strong>and</strong>ard<br />
Class 2: resinous, nutty, green, grassy<br />
0210 Resinous OT Fresh sawdust, resin, cedarwood, pinewood, sprucy, terpenoid<br />
0211 Woody OT Seasoned wood (uncut)<br />
0220 Nutty OT As in brazil nut, hazelnut, `sherry-like'<br />
0221 Walnut OT Fresh (not rancid) walnut<br />
0222 Coconut OT<br />
0223 Beany OT Bean soup (2,4,7-Decatrienal)<br />
0224 Almond OT Marzipan (Benzaldehyde)<br />
0230<br />
0231<br />
0232<br />
Grassy<br />
Freshy cut grass<br />
Straw-like<br />
OT<br />
OT<br />
OT<br />
Green, crushed green leaves, leafy, alfalfa }<br />
Hay-like<br />
cis-3-Hexanol<br />
Class 3: cereal<br />
0310 Grainy OT Raw grain flavour<br />
0311 Husky OT Husk-like, chaff, Glattwasser<br />
0312 Corn grits OT Maize grits, adjuncty<br />
0313 Mealy OT Like flour<br />
0320 Malty OT<br />
0330 Worty OT Fresh wort aroma; use with other terms to describe infected wort<br />
(e.g. 0731 Parsnip/celery)<br />
Class 4: caramelized, roasted<br />
0410 Caramel OT Burnt sugar, toffee-like<br />
0411 Molasses OT Black treacle, treacly<br />
0412 Licorice OT<br />
0420 Burnt OTM Scorched aroma, dry mouth-feel, sharp, acrid taste<br />
0421 Bread crust OTM Charred toast<br />
0422 Roast barley OTM Chocolate malt<br />
0423 Smoky OT<br />
Class 5: phenolic<br />
0500 Phenolic<br />
0501 Tarry OT Pitch, faulty pitching of containers<br />
0502 Bakelite OT<br />
0503 Carbolic OT Phenol, C 6 H 5OH<br />
0504 Chlorophenol OT Trichlorophenol (TCP), hospital-like<br />
0505 Iodoform OT Iodophores, hospital-like, pharmaceutical<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Class 6: soapy, fatty, diacetyl, oily, rancid<br />
0610 Fatty acid OT<br />
0611 Caprylic OT Soapy, fatty, goaty, tallowy (Octanoic acid)<br />
0612 Cheesy OT<br />
Dry, stale cheese, old hops<br />
0613 Isovaleric OT Hydrolytic (Isovaleric acid)<br />
0614 Butyric OT Rancid butter rancidity Butyric acid 3 mg/1<br />
0620 Diacetyl OT Butterscotch, buttermilk Diacetyl, 0.2±0.4 mg/1<br />
0630 Rancid OT<br />
Oxidative rancidity<br />
0631 Rancid oil OTM<br />
0640 Oily OTM<br />
0641 Vegetable oil OTM As in refined vegetable oil<br />
0642 Mineral oil OTM Gasoline (petrol) kerosene (paraffin), machine oil<br />
Class 7: sulphury<br />
}<br />
}<br />
0700 Sulphury OT<br />
0710 Sulphitic OT Sulphur dioxide, striking match, choking, sulphurous-SO 2 (KMS)<br />
0720 Sulphidic OT Rotten egg, sulphury-reduced, sulphurous- RSH<br />
0721 H 2S OT Rotten egg (H 2S)<br />
0722 Mercaptan OT Lower mercaptans, drains, stench (Ethyl mercaptan)<br />
0723 Garlic OT<br />
0724 Lightstruck OT Skunky, sunstruck<br />
0725 Autolysed OT Rotting yeast; see also 0740 Yeasty<br />
0726 Burnt rubber OT Higher mercaptans<br />
0727 Shrimp-like OT Water in which shrimp have been cooked<br />
0730 Cooked OT Mainly dialkyl sulphides, sulphurous-RSR<br />
vegetable<br />
0731 Parsnip/celery OT An effect of wort infection<br />
0732 DMS OT (Dimethyl sulphide) DMS, 100 g/l<br />
0733 Cooked cabbage OT Over-cooked green vegetables<br />
0734 Cooked sweet corn OT Cooked maize, canned sweet corn<br />
0735 Cooked tomato OT Tomato juice (processed), tomato ketchup<br />
0736 Cooked onion OT<br />
0740 Yeasty OT Fresh yeast, flavour of heated thiamine (see also 0725 autolysed)<br />
0741 Meaty OT Broth, cooked meat, meat extract, peptone, yeast broth<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
}
Table 20.11 Continued<br />
Class term First tier Second tier Relevance Comments, synonyms, definitions Reference st<strong>and</strong>ard<br />
Class 8: oxidized, stale, musty<br />
0800 Stale OTM Old beer, overaged, overpasteurized (Heat with air)<br />
0810 Catty OT Blackcurrant leaves, ribes, tomato plants, oxidized beer (p-Methane-8-thiol-3one)<br />
0820 Papery OT Initial stage of staling, bready (stale bread crumb), cardboard, old (5 Methylfurfural,<br />
beer, oxidized 25 mg/1)<br />
0830 Leathery OTM Later stage of staling, often used in conjunction with 0211 Woody<br />
0840 Mouldy OT Cellar-like, leaf mould, woodsy<br />
0841 Earthy OT Actinomycetes, damp soil, freshly dug soil, diatomaceous earth (Geosmin)<br />
0842 Musty OT Fusty<br />
Class 9: sour, acidic<br />
0900 Acidic OT Pungent aroma, sharpness of taste, mineral acid<br />
0910 Acetic OT Vinegar (Acetic acid)<br />
0920 Sour OT Lactic, sour milk: use with 0141 citrus for citrus-sour<br />
Class 10: sweet<br />
1000 Sweet OT Sucrose 7.5 g/1<br />
1001 Honey OT Can occur as effect of beer staling (e.g. odour of stale beer in glass),<br />
oxidized (stale) honey<br />
1002 Jam-like OT May be qualified by subclasses of 0140 Fruity<br />
1003 Vanilla OT Custard powder, vanillin (Vanillin)<br />
1004 Primings OT<br />
1005 Syrupy OTM Clear (golden) syrup<br />
1006 Oversweet OT Sickly sweet, cloying<br />
Class 11: salty<br />
1100 Salty T Sodium chloride, 1.8 g/1<br />
Class 12: bitter<br />
1200 Bitter TAf (Isohumulone)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 20.11 Continued<br />
Class 13: mouth-feel<br />
1310 Alkaline TMAf Flavour imparted by accidental admixture of alkaline detergent (Sodium bicarbonate)<br />
1320 Mouthcoating MAf Creamy, onctueux (Fr)<br />
1330 Metallic OTMAf Iron, rusty water, coins, tinny, inky (Ferrous ammonium<br />
sulphate)<br />
1340 Astringent MAf Mouth puckering, puckery, tannin-like, tart Quercitrin, 240 mg/1*<br />
1341 Drying MAf Unsweet<br />
1350 Powdery OTM O-Dusty cushion, irritating, (with 0310 Grainy) mill room smell<br />
TM-Chalky, particulate, scratchy, silicate-like, siliceous<br />
1360 Carbonation M CO 2 content<br />
1361 Flat M Undercarbonated 60% of normal C0 2<br />
content for the product<br />
1362 Gassy M Overcarbonated 140% of normal C0 2<br />
content for the product<br />
1370 Warming WMAf See 0110 Alcoholic <strong>and</strong> 0111 Spicy<br />
Class 14: fullness<br />
1410 Body OTM Fullness of flavour <strong>and</strong> mouth-feel<br />
1411 Watery TM Thin, seemingly diluted<br />
1412 Characterless OTM Bl<strong>and</strong>, empty, flavourless<br />
1413 Satiating OTM Extra full, filling<br />
1414 Thick TM Viscous, eÂpais (Fr)<br />
* Quercitrin is both astringent <strong>and</strong> bitter<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
3. No terms are duplicated for the same flavour characteristic. (In five cases,<br />
overlapping pairs of chemical name terms <strong>and</strong> generally descriptive terms had to be<br />
permitted. The five pairs are: 0131 isoamyl acetate <strong>and</strong> 0143 banana; 0132 ethyl<br />
hexanoate <strong>and</strong> 0142 apple; 0133 ethyl acetate <strong>and</strong> 0120 solvent like; 0613 isovaleric<br />
<strong>and</strong> 0612 cheesy; <strong>and</strong> 0732 DMS <strong>and</strong> 0734 cooked sweet corn.)<br />
4. The system is compatible with the EBC Thesaurus for the <strong>Brewing</strong> Industry.<br />
5. Subjective terms such as good/bad, young/mature, balanced/unbalanced are not<br />
included.<br />
6. As far as possible the meaning of each term is illustrated with readily available<br />
reference st<strong>and</strong>ards.<br />
The system (Table 20.11) consists of 14 classes given general names to indicate the area<br />
in which any given type of flavour should be sought. Descriptors carry afour-digit<br />
number. Some classes have abroader term (e.g. 0700 sulphury) that serves as acommon<br />
descriptor for all the terms in the class; in other classes asuitable term is not available.<br />
There are three kinds of descriptors: class terms, first-tier terms <strong>and</strong> second-tier terms.<br />
In general the first two contain common terms familiar to most people, <strong>and</strong> together they<br />
provide avocabulary designed to fill nonspecialist needs. The flavour wheel (Fig. 20.22)<br />
is presented to facilitate the location of terms within the system. It is a memory aid <strong>and</strong><br />
not a new system of classification. Despite the diversity of terms, a logical sequence is<br />
obtained in most cases, but certain discontinuities appear, as where 0700 sulphury follows<br />
0640 oily. The second tier of terms, together with the reference st<strong>and</strong>ards, form the<br />
theoretical backbone of the system <strong>and</strong> also serve to define those first-tier terms for which<br />
a reference is not available, e.g., 0220 nutty comprises a group of flavour notes<br />
exemplified by walnut like, coconut like, beany, <strong>and</strong> almond like. The column<br />
`Relevance' shows that most terms may be used to describe sensations of both odour (O)<br />
<strong>and</strong> taste (T). The letters M, W, <strong>and</strong> Af indicate that the terms may be used to describe<br />
mouth-feel effects, warming <strong>and</strong> after flavour. A number of terms that have been used in<br />
the past are given under `Comments, synonyms, definitions' but their use should be<br />
discouraged in favour of the more precise description given in the Table. Thus, 0910<br />
acetic is preferred to `vinegar'. The flavour caused by caprylic <strong>and</strong>/or capric acids should<br />
be referred to as 0611 caprylic. The term 0630 rancid is used only for oxidative rancidity<br />
(carbonyl compounds) <strong>and</strong> is no longer used for a butyric flavour.<br />
Analytica-EBC-ASBC provide a list of 27 compounds recommended for use as flavour<br />
reference st<strong>and</strong>ards together with methods of purification, difference thresholds <strong>and</strong> the<br />
range of values found in beer. In addition they list 15 compounds that may be suitable<br />
flavour reference st<strong>and</strong>ards after further study.<br />
It is recognized that terminology will change with usage <strong>and</strong> the results of research<br />
<strong>and</strong> so the system should be brought up to date every few years. Individual breweries may<br />
well use other terms which help to characterize their beers. Lee et al. (2001) have<br />
presented a revised flavour wheel for use with whiskies.<br />
Brown <strong>and</strong> Clapperton (1978b) examined the terms used to describe ale flavours by<br />
multi-dimensional scaling ± a technique of grouping like characters together <strong>and</strong><br />
arranging these groups relative to each other by their degree of `similarity'. Such<br />
correlations cannot be perfect <strong>and</strong> deviations from the model are expressed as `stress'.<br />
Thus the seven after-flavour terms (sweet, toffee-like, caprylic, burnt, astringent, <strong>and</strong><br />
mouthcoating) can be represented in a two-dimensional model with only 0.3% stress. In<br />
contrast for odour <strong>and</strong> flavour terms a three-dimensional model is required <strong>and</strong> the stress<br />
value is approximately 13%. These terms fall roughly on the surface of a sphere so that<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1350 Powdery<br />
1340 Astringent<br />
1330 Metallic<br />
1320 Mouthcoating<br />
1310 Alkaline<br />
1200 Bitter<br />
1100 Salty<br />
1000 Sweet<br />
0920 Sour<br />
0910 Acetic<br />
0900 Acidic<br />
0840 Mouldy<br />
0830 Leathery<br />
12. Bitter<br />
11. Salty<br />
10. Sweet<br />
9. Sour, acidic<br />
0820 Papery<br />
0810 Catty<br />
1360 Carbonation<br />
1370 Warming<br />
13. Mouthfeel<br />
8. Oxidized,<br />
stale, musty<br />
0800 Stale<br />
0740 Yeasty<br />
Taste<br />
0730 Cooked veg.<br />
1410 Body<br />
14. Fullness<br />
0110 Alcoholic<br />
7. Sulphury<br />
0720 Sulfidic<br />
0710 Sulphitic<br />
0700 Sulphury<br />
0120 Solvent-like<br />
0640 Oily<br />
0130 Estery<br />
1. Aromatic, fragrant,<br />
fruity, floral<br />
Odour<br />
Odour<br />
0140 Fruity<br />
0150 Acetaldehyde<br />
0160 Floral<br />
2. Resinous,<br />
nutty, green, grassy<br />
6. Soapy,<br />
fatty, discetyl,<br />
oily, rancid<br />
3. Cereal<br />
4. Caramelized,<br />
roasted<br />
5. Phenolic<br />
0170 Hoppy<br />
0210 Resinous<br />
0220 Nutty<br />
0230 Grassy<br />
0310 Grainy<br />
0320 Malty<br />
0330 Worty<br />
0410 Caramel<br />
0420 Burnt<br />
0500 Phenolic<br />
0610 Fatty acid<br />
0620 Diacetyl<br />
0630 Rancid<br />
Fig. 20.22 Flavour wheel showing class terms <strong>and</strong> first-tier terms (Analytica-EBC).<br />
Toffee-like<br />
Mouthcoating<br />
Viscous<br />
Estery<br />
Body<br />
Warming<br />
Nutty<br />
H.G. fullness<br />
Spicy<br />
Burnt<br />
Malty<br />
Smooth<br />
Worty<br />
Diacetyl<br />
Lively<br />
Sweet<br />
Cloying<br />
Fruity<br />
Smoky<br />
Bitter<br />
Aldehydic<br />
Oily<br />
Grainy<br />
Caprylic<br />
Cooked<br />
vegetable<br />
Cardboard<br />
Watery<br />
Astringent<br />
Musty<br />
Metallic<br />
Sour<br />
Sulphury<br />
Ribes<br />
Acidic<br />
Rancid<br />
Rubbery<br />
Fig. 20.23 Flavour terms by multi-dimensional scaling (after Brown <strong>and</strong> Clapperton, 1978b).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
those that are close together on the model (Fig. 20.23) have aclose flavour relationship,<br />
e.g. sour <strong>and</strong> acidic, whereas widely different or opposite characteristics such as sweet<br />
<strong>and</strong> bitter or body (full) <strong>and</strong> watery (thin) are further apart or at opposite points on the<br />
surface. The flavours associated with strong ales are seen on the left-h<strong>and</strong> side of the<br />
diagram (Fig. 20.23) while many of those on the right-h<strong>and</strong> side are associated with<br />
oxidative deterioration or staling of beer. Terms in the central section of the diagram,<br />
including diacetyl, caprylic, worty, grainy, lively, bitter, burnt <strong>and</strong> nutty, may be regarded<br />
as belonging to an intermediate category of flavours that are pleasant to some <strong>and</strong><br />
unpleasant to others at their normal levels of perceived intensity in commercial beers<br />
High levels of hop bitter substances, particularly in beers that are fermented to dryness,<br />
can impart astringency as well as bitterness to the flavour so these terms are found close<br />
together on the model.<br />
Brown et al. (1974) used discriminant (cluster) analysis to examine national <strong>and</strong><br />
regional differences in lager beers. In this technique each beer is represented as a point in<br />
multi-dimensional space, the coordinates of which are determined either by the individual<br />
flavour characteristics, determined by profile analysis, or by physicochemical parameters,<br />
determined by chemical analysis. Twenty-seven terms gave significant scores with lagers<br />
<strong>and</strong> the pattern of points in 27-dimensional space is simplified by a computer program to<br />
produce eigenvectors (mathematical devices to convert a pattern of points in multidimensional<br />
space into an equivalent pattern of points in a smaller number of<br />
dimensions). This has the advantage of bringing things down to a level nonmathematicians<br />
can visualize but has the disadvantage that the axes only represent<br />
mathematical abstractions <strong>and</strong> not brewing parameters. Thus, the two-dimensional<br />
pattern of North American, Continental European <strong>and</strong> British lagers gave three discrete<br />
tight clusters of points. When only the 12 highest scoring sensory characteristics were<br />
used the beers still fell into three groups but the clusters were more diffuse (Fig. 20.24).<br />
This result can be interpreted in terms of the perceived differences in bitterness, dimethyl<br />
sulphide (DMS) flavour, <strong>and</strong> palate fullness (OG). Of the flavour terms scored: (i)<br />
dimethyl sulphide <strong>and</strong> cabbagy-vegetable water both relate to the DMS factor; (ii) body,<br />
warming, <strong>and</strong> high gravity fullness, <strong>and</strong> viscous relate to differences in original gravity;<br />
B<br />
A<br />
x<br />
(a) (b)<br />
C<br />
y<br />
o<br />
B<br />
Low OG<br />
<strong>and</strong> low<br />
DMS<br />
content<br />
Low<br />
bitterness<br />
A<br />
High<br />
bitterness<br />
High OG <strong>and</strong><br />
high DMS<br />
content<br />
Fig. 20.24 Discriminant analysis of sensory data on thirty-three lager beers. (a) Result; (b)<br />
Interpretation of result. Code: A = North American beers, B = British beers, <strong>and</strong> C =<br />
Continental European beers (after Brown <strong>and</strong> Clapperton, 1978a).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
C
Table 20.12 Results of cluster analysis of sensory <strong>and</strong> physiochemical data on beer (after<br />
Clapperton, 1979)<br />
Odour items Physiochemical parameters Flavour terms<br />
Original gravity<br />
Isocamyl alcohol Body<br />
2-Methylbutanol Estery 1<br />
Total carbohydrates Fruity<br />
Dextrins Viscous<br />
Potassium<br />
Propanol<br />
Mouth coating<br />
Warming 2<br />
4 Estery Ethyl acetate Spicey<br />
Isoamyl acetate * Watery<br />
Phosphate<br />
12 Ribes Air content<br />
16 Hoppy Dodecanoic acid<br />
Tetradeavoic acid<br />
Total fatty acids<br />
Octanoic acid Caprylic 4<br />
Decanoic acid<br />
Sweet<br />
Present gravity Cloying<br />
Toffee-like 5<br />
* Astringent<br />
Bitterness Bitter 9<br />
Sulphur Metallic 17<br />
* Negatively correlated with other terms <strong>and</strong> parameters in the same cluster<br />
<strong>and</strong> (iii) bitter, drying, <strong>and</strong> possibly smooth (mouth-feel) relate to differences in<br />
bitterness.<br />
Brown <strong>and</strong> Clapperton (1978a) also examined 46 ales (OG 1030±1050) from five<br />
brewing companies by sensory profile analysis <strong>and</strong> by instrumental analysis. The most<br />
important variables in the discriminant analysis were: (i) isoamyl alcohol content<br />
(instrumental), (ii) caprylic flavour (sensory), (iii) sodium content (instrumental); (iv)<br />
meaty aroma (sensory), (v) ethyl acetate content (instrumental), (vi) bitter after-flavour<br />
(sensory), (vii) dextrin content (instrumental), (viii) smoky flavour (sensory), <strong>and</strong> (ix)<br />
DMS content (instrumental). On the basis of these parameters 87% of the ales were<br />
correctly assigned to their brewing company. Similarly, the ales could be assigned to their<br />
gravity b<strong>and</strong> either instrumentally on the basis of alcohol or dextrin content or by sensory<br />
analysis using 13 parameters. The sensory parameters in order of importance were: (i)<br />
body, (ii) aldehyde (odour), (iii) high gravity fullness, (iv) viscous (thick), (v) estery, (vi)<br />
meaty (odour), (vii) cooked vegetable (odour), (viii) rubbery, (ix) caprylic, (x) DMS<br />
(odour), (xi) fruity, (xii) cloying, <strong>and</strong> (xii) sour.<br />
Clapperton (1979) also carried out another cluster analysis using both sensory terms <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
physicochemical parameters <strong>and</strong> the results are given in Table 20.12, where the numbers<br />
refer to the order of formation of the clusters, i.e., the lower the number the better the<br />
correlation. Clusters which only contained sensory terms were omitted. The first two<br />
clusters isolate terms that relate to esters, alcohols <strong>and</strong> original gravity. Although there is a<br />
curious reversal of ester <strong>and</strong> alcohol contents <strong>and</strong> the corresponding flavour effect between<br />
clusters 1 <strong>and</strong> 2, estery (odour) is grouped, as expected, with ethyl <strong>and</strong> isoamyl acetate<br />
content in cluster 4. Apart from cluster 16 the clusters indicate causative relationships<br />
between the physicochemical parameters <strong>and</strong> the corresponding sensory terms.<br />
Principal component analysis is another statistical technique that Clapperton <strong>and</strong><br />
Piggott (1979) applied to the results of profile analysis. Ales <strong>and</strong> lagers were examined<br />
<strong>and</strong> two-dimensional plots of the results using the first two principal components as axes<br />
showed resolution of the ales from the lagers <strong>and</strong> the close proximity of duplicate<br />
samples. Piggott <strong>and</strong> Jardine (1979) used principal component analysis to differentiate<br />
various br<strong>and</strong>s of whisky. Most workers in this field have excluded hedonic expressions<br />
but Moll et al. (1978) used principal component analysis to classify Continental European<br />
beers as good, average or poor on the basis of nine physicochemical parameters: colloidal<br />
stability (7 days at 40 ëC/1 day at 0 ëC), cold sensitivity (24 h at 0 ëC), brightness at 12 ëC,<br />
six months test, the content of -phenylethanol, ethyl caprylate, isoamyl acetate <strong>and</strong><br />
isobutanol <strong>and</strong> foam stability.<br />
Hoff et al. (1978) used headspace analysis to determine the levels of the isoamyl<br />
alcohols, isobutanol, ethyl acetate <strong>and</strong> isoamyl acetate in beers ( -phenylethanol <strong>and</strong> ethyl<br />
caprylate are not sufficiently volatile to be measured by headspace analysis). To compare<br />
two beers, the peak areas on the chromatogram (excluding ethanol <strong>and</strong> the internal<br />
st<strong>and</strong>ard xylene) were expressed as a percentage of the total peak area. A chronologically<br />
updated data base was used to calculate the st<strong>and</strong>ard deviation for each peak <strong>and</strong> a twotailed<br />
t-test was performed on the mean values from duplicate determinations on the two<br />
beers. These results were used to predict the results of triangular taste tests:<br />
1. If none of the peaks is significantly different between samples at 0.005 risk, one<br />
predicts that the tasting panel results will be insignificant at 0.05 risk.<br />
2. If one or more peaks are significantly different between samples at 0.001 risk, one<br />
predicts that the panel results will be significant at a risk of 0.05 or less.<br />
3. If one or more peaks are significantly different between samples at 0.005 risk but<br />
insignificant at 0.001 risk, no prediction is made <strong>and</strong> the sample number increased.<br />
The taste panel found 200 significant results out of the 234 predicted <strong>and</strong> 70<br />
insignificant results out of the 76 predicted <strong>and</strong> it was suggested that tastings could be<br />
reduced by eliminating samples that were similar by headspace analysis. However, it was<br />
acknowledged that differences due to sulphur compounds, staling or certain hop<br />
compounds may go undetected. Nevertheless the headspace/statistical method predicted<br />
<strong>and</strong> the tasting panel found significant differences between beers produced at three<br />
branch plants.<br />
Compared to chemical assays, flavour results from human assessors have been<br />
regarded as unreliable but Hegarty <strong>and</strong> White (1993) have applied two-way analysis of<br />
variance (ANOVA) statistics to the results of a flavour profile panel. This can establish<br />
the degree of variability due to assessor differences, the degree of variability due to<br />
differences between samples <strong>and</strong> the amount of variability that cannot be explained<br />
(`noise'). Results can be expressed as the F-ratio (Fisher ratio) which, if large, shows a<br />
significant difference. Thus, ideally, the F-ratio for the assessors should show a low value<br />
indicating that the score attributed to a given sample does not vary greatly between<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
tasters. This approach can be exp<strong>and</strong>ed to monitor the performance of individual<br />
assessors. The mean score can show whether the assessor is scoring high or low relative<br />
to the rest of the group. The st<strong>and</strong>ard deviation is a measure of scoring consistency <strong>and</strong><br />
the F-ratio indicates whether the assessor can distinguish samples consistently. The<br />
statistics showed that tasters do not perform equally on all characters. The key to reliable<br />
flavour results is on-going performance monitoring <strong>and</strong> on-going flavour training.<br />
20.5 References<br />
ACREE, T. E. <strong>and</strong> TERANISHI, R. (eds) (1993) Flavor Science ± Sensible Principles <strong>and</strong> Techniques.<br />
American Chemical Society, Washington DC. pp. xvi + 352.<br />
AKABAS, M. H., DODD, J. <strong>and</strong> AL-AWQATI, Q. (1988) Science, 242, 1047.<br />
AMERICAN SOCIETY FOR TESTING AND MATERIALS (1979) St<strong>and</strong>ard <strong>practice</strong> for Determination of Odour<br />
<strong>and</strong> Taste Thresholds by Forced Choice Ascending Concentration Series. E 67979. ASTM,<br />
Philadelphia, PA.<br />
AMERINE, M. A., PANGBORN, R. M. <strong>and</strong> ROESSLER, E. B. (1965) Principles of Sensory Evaluation of Food.<br />
Academic Press, New York.<br />
AMOORE, J. E. (1991) in Smell <strong>and</strong> Taste in Health <strong>and</strong> Disease (ed. Getchell, T. V. et al.) p. 655.<br />
BOUDREAU, J. C. (ed.) (1979) Food Taste Chemistry. ACS Symposium No. 115. American Chemical<br />
Society, Washington DC, 262pp.<br />
BRESLIN, P. A. S. (2001) Flavour Fragr. J. 16, 439.<br />
BROWN, D. G. W. <strong>and</strong> CLAPPERTON, J. F. (1978a) J. Inst. <strong>Brewing</strong>, 84, 318.<br />
BROWN, D. G. W. <strong>and</strong> CLAPPERTON, J. F. (1978b) J. Inst. <strong>Brewing</strong>, 84, 324.<br />
BROWN, D. G. W., CLAPPERTON, J. F. <strong>and</strong> DALGLIESH, C. E. (1974) Proc. Annu. Meet. Am. Soc. Brew.<br />
Chem., p. 1.<br />
BROWN, D. G. W., CLAPPERTON, J. F., MEILGAARD, M. C. <strong>and</strong> MOLL, M. (1978) J. Amer. Soc. Brew. Chem.,<br />
36, 73.<br />
CHEVANCE, F., GUYOT-DECLERCK, C., DUPONT, J. <strong>and</strong> COLLIN, S. (2002) J. Agric. Food Chem., 50, 3818.<br />
CLAPPERTON, J. F. (1974) J. Inst. <strong>Brewing</strong>, 80, 164.<br />
CLAPPERTON, J. F. (1976) J. Inst. <strong>Brewing</strong>, 82, 175<br />
CLAPPERTON, J. F. (1978) J. Inst. <strong>Brewing</strong>, 84, 107.<br />
CLAPPERTON, J. F. (1979) in L<strong>and</strong>, D. G. <strong>and</strong> Nursten, H. E. Progress in Flavour Research. Applied<br />
Science Publishers, London, p. 1.<br />
CLAPPERTON, J. F. <strong>and</strong> BROWN, D. L. W. (1978) J. Inst. <strong>Brewing</strong>, 84, 90.<br />
CLAPPERTON, J. F. <strong>and</strong> PIGGOTT, J. R. (1979) J. Inst. <strong>Brewing</strong>, 85, 271.<br />
DALGLIESH, C. E. (1977) Proc. 16th Congr. Eur. Brew. Convn. Amsterdam, p. 623.<br />
DE COOMAN, L., AERTS, G., OVERMEIRE, H. <strong>and</strong> DE KEUKELEIRE, D. (2000) J. Inst. <strong>Brewing</strong>, 106, 169.<br />
DELWICHE, J. F., BULETIC, Z. <strong>and</strong> BRESLIN, P. A. S. (2001) Perception <strong>and</strong> Psychophysics, 63, 761.<br />
DOTY, R. L. (ed.) (1995) H<strong>and</strong>book of Olfaction <strong>and</strong> Gustation. Marcel Dekker, New York.<br />
EVANS, D. J., SCHMEDDING, D. J. M., BRUIJNJE, A., HEIDEMAN, T. <strong>and</strong> KING, B. M. (1999) J. Inst. <strong>Brewing</strong>,<br />
105, 301.<br />
FERNANDUS, A., OOSERAM-KLEIJNGELD, I. <strong>and</strong> RUNNEBOOM, A. J. M. (1970) Tech. Quart. MBBA, 7, 210.<br />
GARDNER, R. J. (1978) J. Pharm. Pharmacol., 30, 351.<br />
GARDNER, R. J. (1979) Tech. Quart. MBAA, 16, 106, 148, 204.<br />
GIENAPP, E. <strong>and</strong> SCHROÈ DER, K. L. (1975) Die Nahrung, 19, 697.<br />
GIJS, L. <strong>and</strong> COLLIN, S. (2002) J. Amer. Soc. Brew. Chem., 60, 68.<br />
GIJS, L., PERPEÁ TE, P., TIMMERMANS, A. <strong>and</strong> COLLIN, S. (2000) J. Agric. Food Chem., 48, 6196.<br />
GIJS, L., CHEVANCE, F., JERKOVIC, V. <strong>and</strong> COLLIN, S. (2002) J. Agric. Food Chem., 50, 5612.<br />
GIVEN, P. <strong>and</strong> PARADES, D. (eds.) (2002) Chemistry of taste ± Mechanisms, Behaviors <strong>and</strong> Mimics. ACS<br />
Symposium No. 825. American Chemical Society, Washington, DC.<br />
GOIRIS, K., DE RIDDER, M., DE ROUCK, G., BOEYKENS, A. VAN OPSTAELE, F., AERTS, G., DE COOMAN, L. <strong>and</strong><br />
DE KEUKELEIRE, D (2002) J. Inst. <strong>Brewing</strong>, 108, 86.<br />
GOODE, J. (2003) Wine Magazine, April, 2003, p. 44.<br />
GOODENOUGH, P. W. (1998) Int. J. Food Sci., 33, 63.<br />
GRAY, P. P. <strong>and</strong> STONE, I. (1939) J. Inst. <strong>Brewing</strong>, 45, 253.<br />
GUADAGNI, D. G. (1970) quoted by Teranishi, R. in Ohloff, G. <strong>and</strong> Thomas, A. F. (eds) (1971) Gustation<br />
<strong>and</strong> Olfaction Academic Press, New York, p. 170<br />
HEGARTY, P. K. <strong>and</strong> WHITE, F. H. (1993) Proc. 24th Congr. Eur. Brew. Convn. Oslo, p. 429<br />
HEGARTY, P., CHILVER, J. <strong>and</strong> THREAPLETON, L. (2001) Proc. 28th Congr. Eur. Brew. Convn. Budapest<br />
Paper 89.<br />
HOFF, J. T., CHICOYE, E., HERWIG, W. C. <strong>and</strong> HELBERT, J. R. (1978) in Charalambous, G. (ed.) (1978)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Analysis of Foods <strong>and</strong> Beverages Headspace Techniques, Academic Press, New York, p. 187.<br />
HOUGH, J. S. (1990) in An Introduction to <strong>Brewing</strong> Science <strong>and</strong> Technology. Series II, Volume 3. Quality.<br />
Institute of <strong>Brewing</strong>, London.<br />
HUGHES, P. S. <strong>and</strong> BOLSHAW, L. H. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, p. 151.<br />
INSTITUTE OF BREWING (1995) Sensory Analysis Manual, 44 pp.<br />
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION (1983) International St<strong>and</strong>ard 6658. Sensory<br />
Analysis, Methodology, General guidance. ISO, Paris.<br />
JONES, F. N. (1956) Amer. J. Psychology, 69, 672.<br />
KANEDA, H., KANO, Y., KOSHINO, S. <strong>and</strong> OHYA-NISHIGUCHI, H. (1992) J. Agric. Food Chem., 40, 2102.<br />
KANEDA, H., SHINOTZUKA, K., KOBAYAKAWA, T., SAITO, S. <strong>and</strong> OKAHATA, Y. (2001) J. Amer. Soc. Brew.<br />
Chem., 49, 167.<br />
KANEDA, H., WATARI, J., TAKASHIO, M. <strong>and</strong> OKAHATA, Y. (2003) J. Inst. <strong>Brewing</strong>, 109, 27.<br />
KIER, L. B. <strong>and</strong> HALL, L. H. (1976) Molecular Connectivity in Chemistry <strong>and</strong> Drug Research. Academic<br />
Press, New York.<br />
LACAN, F., SOULET, S., ARNAUDINAUD, V., NAY, B., VERGEÂ , S., CASTAGNINO, C., DELAUNAY, J.-C., CHEÁ ZE, C.<br />
<strong>and</strong> VERCAUTEREN, J. (2000) Cerevisia, 25 (4), 35.<br />
LANGSTAFF, S. A. <strong>and</strong> LEWIS, M. J. (1993) J. Inst. <strong>Brewing</strong>, 99, 31.<br />
LAWLESS, H. <strong>and</strong> HEYMANN, H. (1999) Sensory Analysis of Foods: Principles <strong>and</strong> Practice. Kluwer/<br />
Plenum, New York.<br />
LEE, K.-Y. M., PATTERSON, A., PIGGOTT, J. R. <strong>and</strong> RICHARDSON, G. D. (2001) J. Inst. <strong>Brewing</strong>, 107, 287.<br />
LERMUSIEAU, G., NOEÈ L, S., LIEÂ GEIS, C. <strong>and</strong> COLLIN, S. (1999) J. Amer. Soc. Brew. Chem., 57, 29.<br />
LEWIS, M. J., PANGBORN, R. M. <strong>and</strong> FUJII-YAMASHITA, J. (1980) Proc. 16th Conv. Australian <strong>and</strong> New<br />
Zeal<strong>and</strong> section of the Institue of <strong>Brewing</strong>, Sydney, p. 165.<br />
MACKIE, A. E. <strong>and</strong> SLAUGHTER, J. C. (2002) J. Inst. <strong>Brewing</strong>, 108, 336.<br />
MALEK, D. M., SCHMIDT, D. J. <strong>and</strong> MUNROE, J. H. J. (1982) J. Amer. Soc. Brew. Chem., 40, 133.<br />
MEILGAARD, M. C. (1975) Tech. Quart. MBAA, 12, 107, 151.<br />
MEILGAARD, M. C. <strong>and</strong> REID, D. S. (1979) in L<strong>and</strong>, D. G. <strong>and</strong> Nursten, H. E. (eds) Progress in Flavour<br />
Research. Applied Science Publishers, London, p. 67.<br />
MEILGAARD, M. C., CIVILLE, G. V. <strong>and</strong> CARR, B. T. (1987) Sensory Evaluation Techniques. CRC Press,<br />
Boca Raton FL.<br />
MOLL, M., VINH, T. <strong>and</strong> FLAYEUX, R. (1978) in Charalambous, G. (ed.) Flavours of Foods <strong>and</strong> Beverages ±<br />
Chemistry <strong>and</strong> Technology. Academic Press, New York, p. 329.<br />
MONTMAYEUR, J. P. <strong>and</strong> MATSUNAMI, H. (2002) Current Opinion in Neurobiology, 12, 366.<br />
NIEMAN, C. (1960) quoted in Amerine, M. A., Pangborn, R.M. <strong>and</strong> Roessler, E. B. (1965) Principles of<br />
Sensory Evaluation of Food. Academic Press, New York, p. 95.<br />
NOEÈ L, S., METAIS, N., BONTE, S., BOPART, E., PELADAN, F., DUPIRE, S. <strong>and</strong> COLLIN, S. (1999) J. Inst. <strong>Brewing</strong>,<br />
105, 269.<br />
OHLOFF, G. (1994) Scent <strong>and</strong> Fragrances ± The Fascination of Odours <strong>and</strong> their Chemical Perspectives.<br />
Springer Verlag, Berlin, pp. xii + 238. Translated from the German Riechstoffe und Geruchssinn<br />
(1990) by Pickenhagen, W. <strong>and</strong> Lawrence, B. M.<br />
PANGBORN, R. M. (1959) Amer. J. Clin. Nutrition, 7, 280.<br />
PANGBORN, R. M. (1980) in Koivistoinen, P. <strong>and</strong> HyvoÈnen. L. Carbohydrate Sweeteners in Food <strong>and</strong><br />
Nutrition, Academic Press, New York, p. 87.<br />
PERPEÁ TE, P. <strong>and</strong> COLLIN, S. (1999) J. Agric. Food Chem., 47, 2374.<br />
PIGGOTT, J. R. (ed.) (1988) Sensory Analysis of Foods. 2nd edn, pp. x + 422. Elsevier Applied Science,<br />
London.<br />
PIGGOTT, J. R. <strong>and</strong> JARDINE, S. P. (1979) J. Inst. <strong>Brewing</strong>, 85, 82.<br />
RICHTER, C. P. <strong>and</strong> MACLEAN, A. (1939) Amer. J. Physiol., 126, 1.<br />
ROESSLER, E. B., PANGBORN, R. M., SIDEL, J. L. <strong>and</strong> STONE, H. J. (1978) Food Science, 43, 940.<br />
ROUSEFF, R. L. (ed.) (1990) Bitterness in Foods <strong>and</strong> Beverages. Elsevier, Amsterdam, pp xviii + 356.<br />
SMYTHE, J. E., O'MAHONY, M. A. <strong>and</strong> BAMFORTH, C. W. (2002) J. Inst. <strong>Brewing</strong>, 108, 37.<br />
SWAIN, T. <strong>and</strong> CASEY, J. C. unpublished results quoted in Reynolds, T. M (1963) Adv. Food Research, 12,<br />
1.<br />
TEETER, J. H. <strong>and</strong> GOLD, G. H. (1988) Nature, 331, 298.<br />
TORLINE, P. A., DERCKSEN, A. W., AXCELL, B. C. <strong>and</strong> JOHNSTONE, W. (1999) Proc. 7th Conv. Inst. <strong>Brewing</strong><br />
Africa Section, Nairobi, p. 50.<br />
TRESSL, R., BAHRI, D. <strong>and</strong> KOSSA, M. (1980) in Charalambous, G (ed.) The Analysis <strong>and</strong> Control of Less<br />
Desirable Flavours in Foods <strong>and</strong> Beverages. Academic Press, New York, p. 293.<br />
UCHIDA, M. <strong>and</strong> ONO, M. (1999) J. Amer. Soc. Brew. Chem., 57, 145.<br />
VAN DER HEIJDEN, A. (1993) in Acree, T. A. <strong>and</strong> Teranishi, R. (eds) Flavor Science ± Sensible Principles<br />
<strong>and</strong> Techniques. American Chemical Society, Washington, DC, p. 67.<br />
VAN OEVELEN, D., DE L'ESCAILLE, F. <strong>and</strong> VERACHTERT, H. (1976) J. Inst. <strong>Brewing</strong>, 82, 322.<br />
WEISS, A., SCHOÈ NBERGER, CH., MITTER, W., BIENDL, M., BACK, W. <strong>and</strong> KROTTENTHALER, M. (2002) J. Inst.<br />
<strong>Brewing</strong>, 108, 236.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
21<br />
Packaging<br />
21.1 Introduction<br />
Beermustbepackagedbeforeitissold.Toensurethebestpossiblequalityoftheproduct,<br />
packaging must be carried out with skill <strong>and</strong> care. Only if packaging is effectively<br />
performedwilltheproductbeacceptable.Beercanbeputintoanumberofpackages.The<br />
most important world-wide is the bottle. Bottles are of two types: returnable <strong>and</strong> nonreturnable.<br />
The most used isthe returnable bottle but indeveloped markets in Europe <strong>and</strong><br />
theUSAthenon-returnablebottleisprevalent.Beerisalsofilledintocans,kegs<strong>and</strong>casks.<br />
Usually adistinction is made between draught beer, i.e., in kegs or casks <strong>and</strong> `small-pack<br />
beer' in bottles <strong>and</strong> cans. There are differences between countries in the relative<br />
proportionsof different packages(Table21.1). The UK <strong>and</strong> Irel<strong>and</strong> are unusualin having<br />
most of their beer on draught. The UK is further unusual in that of the 64% of its beer sold<br />
on draught in 1998, 11% was conditioned in the cask <strong>and</strong> not filtered in the brewery. Cask<br />
conditioned beer dem<strong>and</strong>s very different packaging from keg beer.<br />
In the mature beer markets of Western Europe, Australia <strong>and</strong> the USA sales of beer in<br />
recent years have shown little total growth. In the UK there has been a decline of about<br />
1% per annum for the last ten years. This has resulted in product differentiation efforts<br />
being focused on packaging, particularly on small pack beers. This has coincided with an<br />
increase in beer consumed at home with the consequent domination of this trade by<br />
supermarket groups.<br />
Packaging is influenced by environmental issues, which are stronger in some countries<br />
than others. In some cases a revival in the use of returnable glass bottles <strong>and</strong> the<br />
outlawing of selling beer in cans has occurred. Most countries now have packaging<br />
legislation, which seeks to control the use of packaging material <strong>and</strong> to reduce waste.<br />
This sometimes leads to conflict with marketing where packaging plays such a huge role<br />
in product attractiveness. Packaging is the most labour-intensive part of the brewing<br />
process. The machinery for packaging beer has become progressively more complex with<br />
the object of reducing labour costs <strong>and</strong> preserving product quality. Capital employed in<br />
packaging is usually the highest of the brewing operations. The efficiency of operation of<br />
packaging machinery is of critical importance to a profitable brewery.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 21.1 Relative proportion of beer sold in draught <strong>and</strong> small-pack form for important beerproducing<br />
countries in 2000 (BLRA, 2002)<br />
Country Production ('000hl) Draught sales (%) Small-pack sales (%)<br />
USA 233,521 9 91<br />
China 220,485 5 95<br />
Germany 110,429 19 81<br />
Brazil 82,600 1 99<br />
Japan 71,727 16 84<br />
UK 55,729 62 38<br />
The Netherl<strong>and</strong>s 25,072 30 70<br />
Czech Republic 17,924 46 54<br />
Australia 17,326 24 76<br />
Irel<strong>and</strong> 8,710 78 22<br />
Denmark 7,460 11 89<br />
It is essential to keep records of packaging operations. These relate to the strength <strong>and</strong><br />
type of the beer packaged <strong>and</strong> to the volume of the beer. Everywhere there is legislation<br />
governing the contents of the beer in the package for sale. This may relate to average or<br />
to aguaranteed minimum content. The records will be audited by officials. In most<br />
countries atax (excise duty) is taken relating to the strength of the beer. The packaging<br />
department or warehouse keeps the records on which this is based (Chapter 22).<br />
Packaging is thus of fundamental importance in the supply of beer <strong>and</strong> is pivotal in<br />
ensuring the customer is satisfied in terms of quality, quantity <strong>and</strong> legality. The<br />
preparation of beer for packaging was discussed in Chapter 15. The packaging options<br />
available, particularly for bottles <strong>and</strong> cans, are now numerous <strong>and</strong> involve different types<br />
of multi-pack presentation using cardboard <strong>and</strong> plastic. Br<strong>and</strong>s can establish an identity<br />
based on the package alone <strong>and</strong> this is frequently as important as the identity created by<br />
the taste of the beer. This chapter deals with the underlying principles of successful<br />
modern packaging operations.<br />
21.2 General overview of packaging operations<br />
Apackaging line is aseries of machines designed to fill containers with beer <strong>and</strong> present<br />
those containers (packages) to the warehouse. The detailed design of the line will depend<br />
on the type of package (bottle, can, keg, or cask), the required rate of packaging <strong>and</strong> the<br />
types of beer to be packaged. There will also be machines to deal with any secondary<br />
packaging required which is usually specified by the customer.<br />
Modernsmall-packbeerfillersoperateatveryhighrates,bottlingatover1000bottles/<br />
min. <strong>and</strong> canning at 2000 cans/min. is common. As there is little storage space in<br />
breweries for empty cans <strong>and</strong> bottles aconstant stream of bottles or cans to the site is<br />
needed. This could mean around 30 vehicles/24hcarrying 26 pallets of empty containers<br />
arriving at the brewery. Manufacturers of packages are frequently located near to the<br />
packaging plant. The pressure to supply is not so intense with returnable bottles <strong>and</strong> kegs<br />
butthe recovery of empties from the trade must be arranged. These logistics (Chapter 22)<br />
are very important in the management of packaging.<br />
Twoseparateflowsmustbedealtwithinanypackagingplant:theflowofthebeer<strong>and</strong><br />
the flow of the containers both empty <strong>and</strong> full. Thus the mechanical engineering of<br />
machinery with large moving parts <strong>and</strong> its proximity to aperishable foodstuff must be<br />
considered. The h<strong>and</strong>ling of the container as well as the h<strong>and</strong>ling of the beer must be<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
optimized. Three main aspects characterize all successful packaging operations:<br />
· Preventing air getting into the beer is essential. All the precautions followed in<br />
producing bright beer must be maintained. Probably the final product specification for<br />
dissolved oxygen in the beer will be < 0.2 mg/l or, in some cases, < 0.1 mg/l, so during<br />
filling operations the pick-up of oxygen must not exceed 0.02±0.03 mg/l. If this<br />
oxygen level is not achieved serious flavour deterioration will result. This control of<br />
oxygen is therefore a major feature of good packaging. Control of oxygen on filling is<br />
not important with casks. Cask beer contains live yeast (1±4 million cells/ml) for<br />
secondary fermentation <strong>and</strong> any oxygen present in the beer is rapidly scavenged by the<br />
yeast.<br />
· The temperature of the beer after conditioning <strong>and</strong> on filling the bright beer tank is<br />
likely to be 1 to 0 ëC (30±32 ëF) <strong>and</strong> the carbon dioxide content may be 2.1±2.7<br />
volumes depending on whether it is destined for keg or small pack. The pressure on<br />
the beer must be maintained <strong>and</strong> the temperature rise controlled to < 2 ëC (3.5 ëF) to<br />
keep carbon dioxide in solution. Carbon dioxide loss is a serious problem, which can<br />
disrupt beer supply as beer is held in the warehouse pending re-processing. This is<br />
expensive <strong>and</strong> potentially damaging to customer service.<br />
· The final major factor common to all filling plants is cleanliness. All the plant, not just<br />
that in contact with the beer, must be regularly <strong>and</strong> thoroughly cleaned.<br />
21.3 Bottling<br />
Worldwide most beer is drunk from bottles, either returnable or non-returnable. The<br />
filling of these bottles is essentially the same. The difference in equipment needed relates<br />
to the h<strong>and</strong>ling of the used <strong>and</strong> returned empty bottles <strong>and</strong> washing them prior to filling.<br />
Many of the principles of successful bottling apply equally to canning, <strong>and</strong> to some<br />
extent, to kegging. Details will be discussed in this section on bottling <strong>and</strong> differences<br />
highlighted in the sections on canning <strong>and</strong> kegging.<br />
A successful bottling line should allow the brewer to:<br />
· maintain the dissolved oxygen level in the beer to at least < 0.2 mg/l, although there<br />
are now reports of plant able to meet a specification of < 0.05 mg/l (Parsons, 2000)<br />
· ensure the beer is supplied to the customer containing no viable micro-organisms<br />
· operate with the minimum number of stoppages to keep losses to < 1.5% of the total<br />
brewery loss <strong>and</strong> lower the headspace air content to < 2ml/l <strong>and</strong> provide the highest<br />
efficiency.<br />
The bottling line consists of a series of machines <strong>and</strong> processes:<br />
· depalletizer<br />
· decrater<br />
· washer<br />
· empty bottle inspection<br />
· flash pasteurization or sterile filtration<br />
· filler<br />
· crowner<br />
· tunnel pasteurization<br />
· full bottle inspection<br />
· labeller<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
· crater<br />
· palletizer<br />
· cleaning<br />
If the beer is flash pasteurized or sterile filtered then the subsequent operations must be<br />
aseptic. The flow of the beer relates to filling <strong>and</strong>, if in use, sterile filling or flash<br />
pasteurization. All the other machines are associated with the flow of containers both full<br />
<strong>and</strong> empty. Normally there is asystem for in-place cleaning (CIP) that is naturally more<br />
rigorouswhensterilefillingisbeingused.Themostimportant machineisthefiller<strong>and</strong>to<br />
satisfy the criteria above this must operate with the minimum of stops <strong>and</strong> on agiven run<br />
of product, operate continuously. Storage capacities before <strong>and</strong> after the filler must be<br />
designed to allow continuous filling. This can take up alot of space. If the efficiency of<br />
thefilleristakenas100%thenthemachinesoneithersideassociatedwithcontainerflow<br />
must have operational capacities of 110±140% to ensure uninterrupted filling.<br />
21.3.1 Managing the bottle flow<br />
The flow of containers has to be managed before <strong>and</strong> after filling <strong>and</strong> labelling. We are<br />
concernedwith:depalletizing;decrating; washingorrinsing;emptybottleinspection;full<br />
bottle inspection; labelling; crating or other secondary packaging; palletizing. Packaging<br />
is labour intensive. The numbers of people employed have been reduced by automation<br />
<strong>and</strong> by careful planning of the layout of the machinery. This must take account of the<br />
organization of the warehouse for the receipt of full <strong>and</strong> empty goods. The number of<br />
work stations on the packaging line should be minimized <strong>and</strong> hence the numbers<br />
employed in the packaging department should be reduced. The machines involved in<br />
palletizing <strong>and</strong> crating of full containers <strong>and</strong> depalletizing <strong>and</strong> decrating of empty<br />
containers are similar in principle. To reduce labour involvement they must be located<br />
together to form one work station (Fig. 21.1). This type of organization has allowed a<br />
productivity of 0.2 million hl/year/person in a Japanese bottling hall operating at 72,000<br />
bottles/h (Yokoi et al., 1991). Improvements on this figure are now being achieved.<br />
Bottling of returnable bottles involves receiving crates of dirty bottles on pallets<br />
separating the crates from the pallets <strong>and</strong> the bottles from the crates <strong>and</strong> then washing the<br />
empty bottles. Bottling into non-returnable bottles involves receiving new glass bottles,<br />
which only require rinsing not the thorough cleaning associated with washing dirty<br />
bottles.<br />
Depalletizing <strong>and</strong> palletizing<br />
Palletizers <strong>and</strong> depalletizers are normally closely situated <strong>and</strong> operated by one man. The<br />
machines must interface with forklift trucks in the warehouse. The efficient operation of<br />
these machines is critical to overall efficiency of the line. They are normally capable of<br />
working at rates of 1.4 filling rate (filling ˆ 1). Dirty returnable bottles are usually<br />
received at the brewery in crates. New glass bottles for the non-returnable trade are<br />
received on cardboard trays. These crates or trays are most frequently formed onto<br />
pallets, which can be wooden in which case there are layers of crates built onto the pallet<br />
or there can be plastic spacer boards separating layers of crates or trays. Plastic spacer<br />
boards are less robust than wooden pallets but are lighter <strong>and</strong> take up less space.<br />
The depalletizer removes the crates or trays <strong>and</strong> presents them to the decrating<br />
machine. This is achieved by a lifting device fixed to a frame holding a loading head. The<br />
loading head loads or unloads the whole layer at once. Lifting heads can be complex <strong>and</strong><br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
New bottles<br />
Returned<br />
bottles<br />
Filled bottles<br />
in warehouse<br />
Depalletizer Decrater Sorter Washer<br />
Repalletizer<br />
Recrater<br />
Labeller<br />
Check<br />
fill<br />
Inspector<br />
can use clamps or hooks to effect the lift. A mixture of pneumatic or hydraulic rams <strong>and</strong><br />
electric motors provides motive force. Position sensing uses micro-switches or photoelectric<br />
cells. The machine is complex <strong>and</strong> requires regular maintenance to ensure<br />
efficient operation. These machines must be working to design criteria to ensure overall<br />
line efficiency. Removal of the spacer board can be effected mechanically but these<br />
machines have to locate the position accurately <strong>and</strong> repetitively <strong>and</strong> this does not always<br />
happen resulting in much frustration. It is often decided, therefore, to remove the spacer<br />
board or the pallet manually. Obviously the machines involved in palletizing <strong>and</strong><br />
depalletizing are identical but operating in the alternative mode.<br />
Decrating <strong>and</strong> crating<br />
Having disassembled the crate or tray the bottles must now be removed <strong>and</strong> presented to<br />
the next machine in the line, the washer. The crating operation at the other end of the line<br />
will need a machine but depending on the detail of the type of package it may operate<br />
differently from the decrater. Removal or packing of bottles is effected by `gripper heads'<br />
mounted on a frame. For efficiency these machines should function at 1.25 filling<br />
(filling ˆ 1). The machines can operate in a batch format or continuously. The choice<br />
should be made with reference to the efficiency obtained. These machines have many<br />
moving parts <strong>and</strong> need considerable maintenance. Machines that work continuously are<br />
less widely used in breweries because different bottle sizes require setting changes. In<br />
continuous machines the gripper heads run along a track in a synchronous movement <strong>and</strong><br />
are constantly lowered or raised to deal with the constantly moving crates. In the batch<br />
process the bottles are removed or placed in the crate in a discrete step, as the crate is<br />
stationary. Either system will work effectively.<br />
Secondary packaging<br />
Returnable bottles are generally sold for consumption in bars <strong>and</strong> other licensed premises.<br />
As such the bottles are supplied in crates. The crate is not displayed to the customer <strong>and</strong><br />
therefore has no marketing significance. Non-returnable bottles are sold for consumption<br />
at home. These are frequently bought through supermarkets where there is scope for<br />
elaborate secondary packaging using cardboard containers, which can display the br<strong>and</strong><br />
logo <strong>and</strong> colours to the purchaser. Secondary packaging is arranged to present the bottles<br />
in a variety of `multi-packs' to satisfy the customer <strong>and</strong> the marketing department. Thus<br />
bottles can be packed in sleeves of 1 4, 2 2, 2 3, or 2 5 combinations (Parsons,<br />
2000; Wainwright, 1999). These clusters can then be packed into final packages of 8, 12,<br />
Filler<br />
Crowner<br />
Pasteurizer<br />
Fig. 21.1 Arrangement of equipment in a bottling hall (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
15, 18, 20 or 24 bottles. Corrugated or plain cardboard is used <strong>and</strong> shrink-wrapping can<br />
beaddedforfinalprotection.Thepossibilitiesofmulti-packingarealmostendless.Itisin<br />
thisareaofthebusinessthatmostexpenditureonresearch <strong>and</strong>developmentisnowmade.<br />
This is avery important part of the fight for br<strong>and</strong> supremacy.<br />
Washing<br />
This is acritical part of returnable bottling. Bottles return to breweries in various states,<br />
they will be dirty <strong>and</strong> will still have labels attached. The labels may be made of paper or<br />
metallized foils, which may contain tin or aluminium. Bottle cleaning must remove all<br />
thelabels,<strong>and</strong>clean<strong>and</strong>sterilizethebottles,whicharethenpresentedtothefiller.Thisis<br />
an aggressive operation <strong>and</strong> over time the bottles will become scuffed <strong>and</strong> unattractive.<br />
This has been aproblem in developed markets where the appearance of the package is so<br />
important. This has led to the rise in popularity of the one-trip non-returnable bottle,<br />
which can be decorated to avery high st<strong>and</strong>ard. This type of package, unless re-cycled, is<br />
less environmentally acceptable than the returnable bottle.<br />
Important factors in all bottle washers are:<br />
· soaking to remove dirt <strong>and</strong> labels<br />
· jetting to rinse<br />
· temperature<br />
· strength of detergent, which is normally alkaline.<br />
The total time of the operation is normally 10±15 minutes. The sequence of operations<br />
varieswithdifferentwashers.Atypicalsystemistosoakthedirtybottlesinhotwater<strong>and</strong><br />
then pass them through ahot caustic soda solution. Bottles are then successively rinsed<br />
with hot caustic solution, hot water, <strong>and</strong> finally cold fresh water. Bottles are transported<br />
through the machine in rows of perhaps 50±70. The bottles are assembled in lines so that<br />
they cannot fall over. Efforts are made to keep the noise as low as possible (
Fresh<br />
water<br />
Bottles out<br />
Bottles in<br />
Sewer<br />
Sewer<br />
Crate Sewer Sieve cleaning<br />
washer<br />
mechanism<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
m<br />
2<br />
1<br />
8<br />
7 6 5 4<br />
3<br />
Sewer<br />
Fig. 21.2 Bottle-washing machine (by courtesy of Krones).<br />
To<br />
sedimentation<br />
tank<br />
1<br />
2<br />
3<br />
4<br />
5<br />
6<br />
7<br />
8<br />
Pre-soak<br />
Pre-jetting<br />
Main caustic<br />
Post caustic<br />
Warm water 1<br />
Warm water 2<br />
Cold water<br />
Fresh water
must be used to prevent the deposit of calcium carbonate scale. Acid scale preventing<br />
additives <strong>and</strong> chelating agents such as gluconates are added to the rinsing zones to<br />
keep calcium in solution <strong>and</strong> phosphates are added as wetting agents. Aluminium in<br />
some foil labels releases hydrogen gas from the caustic solution. This must be vented<br />
from the washer. About 90% of the caustic solution can be used again without<br />
treatment as it is only lightly contaminated. However, about 10% of the solution is<br />
heavily contaminated with paper, colloids, colour pigments, label adhesives, metal<br />
salts <strong>and</strong> oils. The solids are usually removed by settling in an insulated tank. If<br />
bottling is being performed on a two-shift basis (up to 16 h/day) this settling can be<br />
carried out overnight. If settling is carefully performed then the detergent can last for<br />
very long times in the bottle washer.<br />
Caustic bottle-washing solutions are slowly corrosive to glass bottles. This causes<br />
dissolution of soda lime glass by etching. Scuffing also occurs by the physical abrasion of<br />
the bottles as they rub together on the line, causing visible wear rings on the shoulder <strong>and</strong><br />
base of each bottle. This results in a population of etched, scuffed bottles, which cannot<br />
display the beer to its best effect. This also applies to the ceramic labels, which are<br />
sometimes applied directly to the glass. Detergents have been formulated to reduce the<br />
etching <strong>and</strong> so to enhance the life <strong>and</strong> appearance of bottles (Rouillard, 1999; Rouillard<br />
<strong>and</strong> Howell, 1999). It is anticipated that reduced etching will result in reduced scuffing,<br />
as the glass is stronger.<br />
EDTA <strong>and</strong> phosphates present in caustic bottle-washing solutions accelerate corrosion<br />
of the glass (Rouillard <strong>and</strong> Howell, 1999). A corrosion inhibitor (Divobrite Integra) has<br />
been shown to reduce etching to an extent that the appearance of a bottle after 30 trips<br />
was equivalent to that after 15 trips following the use of a conventional bottle-washing<br />
detergent containing EDTA <strong>and</strong> phosphate. This work is important because the returnable<br />
bottle is environmentally friendly but its use is limited by the intense dem<strong>and</strong> for a<br />
`perfect' package, which the returnable bottle cannot be.<br />
Rinsing<br />
Retailing of small pack beer is intensely competitive. Marketing departments have sought<br />
to gain competitive advantage for their beers with extreme differentiation of the package<br />
(see `secondary packaging' above). The non-returnable one-trip bottle is well suited for<br />
this. The br<strong>and</strong> image can be protected <strong>and</strong> enhanced by a `perfect' package. Much of this<br />
beer is sold in large retail chains for consumption at home. Recognition of the product on<br />
the supermarket shelf is vital to success. Non-returnable bottles are displayed in a wide<br />
variety of attractive secondary packaging for this purpose.<br />
Non-returnable bottles are delivered new to the brewery. They are normally rinsed by<br />
spraying inside <strong>and</strong> outside several times. For rinsing the bottles are turned upside down<br />
<strong>and</strong> then returned to the upright position for filling. Rinsers are now almost integral with<br />
the filling machine. The process is designed to wet the bottles prior to filling <strong>and</strong> to<br />
ensure sterility by killing micro-organisms. Steam is usually jetted into the bottle for this<br />
purpose followed by a purge of sterile air. This results in 0.1 to 0.2 ml of condensate<br />
remaining in the bottle.<br />
Empty bottle inspection<br />
A bottler of beer must demonstrate due diligence in providing a quality <strong>and</strong> wholesome<br />
product. He must also run his bottling plant at the highest efficiency possible consistent<br />
with supplying that quality product. The bottles presented to the filler must be `fit to fill'.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
The bottles must not contain any foreign bodies or in the case of returnable bottles, any<br />
residual caustic solution from the washer. Defective bottles with damaged necks must be<br />
sorted <strong>and</strong> removed. They may be fillable but could cause damage to the filling machine<br />
or result in the beer being presented in an unacceptable package. Inspection of empty<br />
bottles prior to filling is therefore an important process. Originally this inspection was<br />
done by eye but with the very high rate of modern bottle fillers (> 1,000 bottles/min) the<br />
process is now performed electronically. Frequently used systems (Kunze, 1999) employ<br />
very high speed cameras (picture-taking speeds of 1/250,000 s) to look for defects.<br />
Examinations are made of internal <strong>and</strong> external sidewalls <strong>and</strong> the base <strong>and</strong> neck of the<br />
bottle. Residual liquid is detected by infra-red radiation from beneath the bottle.<br />
Defective bottles are rejected. For returnable bottles this can mean rejection rates of 2%<br />
of the total bottles returned. In an ageing bottle population this rate could be higher.<br />
Full bottle inspection<br />
There are two reasons for inspecting full bottles: to check the volume of beer in the bottle,<br />
<strong>and</strong> to check for foreign particles. These processes are necessary no matter how careful is<br />
the empty bottle inspection <strong>and</strong> the filling. Usually there is a legal requirement to<br />
guarantee minimum or average contents of the bottle. Inspection systems normally<br />
involve passing a beam of white light or radiation through the bottle at a defined level.<br />
Bottles not meeting the fill level required are rejected. This process is often repeated after<br />
tunnel pasteurization. Off-line checks are also required <strong>and</strong> records must be kept for<br />
inspection by government agents.<br />
Some beers can now be described as global br<strong>and</strong>s. The br<strong>and</strong> can be brewed <strong>and</strong><br />
packaged in many countries. In this situation the quality <strong>and</strong> consistency of the beer is vital<br />
for continuing success. Sometimes it is necessary to introduce on-line checks for foreign<br />
particles including glass fragments in the bottled beer (L<strong>and</strong>man, 1999) to further safeguard<br />
product integrity <strong>and</strong> wholesomeness. The principle of one machine is to spin the bottle thus<br />
suspending any foreign bodies then quickly to stop the bottle leaving the beer <strong>and</strong> any<br />
unwanted particles spinning. The spinning beer is then examined optically using a<br />
computerized video camera. If differences between consecutive images are found,<br />
indicating contaminating particles, the bottle is rejected. The bottle inspecting device is a<br />
rotating carousel holding 36 bottles. Each bottle has a dedicated camera that moves with the<br />
bottle <strong>and</strong> takes multiple pictures as the carousel rotates. The computer detects moving<br />
particles against a stationary bottle. Particles of below 1 mm in size can be detected. This<br />
type of full bottle inspection provides the consumer with near absolute protection against<br />
foreign bodies <strong>and</strong> the brewer is provided with further enhancement of br<strong>and</strong> image.<br />
Labelling<br />
Managing the flow of bottles is needed before <strong>and</strong> after filling. Managing the flow of beer<br />
is concerned with the filling <strong>and</strong> closing operation <strong>and</strong> rendering of the beer free from<br />
micro-organisms. Labelling of the bottle can be considered as part of the management of<br />
the bottle flow. In the sequence of operations labelling follows full bottle inspection,<br />
which may itself have been preceded by tunnel pasteurization. Labelling is of major<br />
significance in the presentation of the beer br<strong>and</strong>. The label not only tells the drinker what<br />
the beer is but also conveys an image associated with advertising that adds to the overall<br />
appeal of the br<strong>and</strong>. This is particularly the case with international beer br<strong>and</strong>s where<br />
instant identity of the br<strong>and</strong> in different countries is important. The application <strong>and</strong> the<br />
quality of the label must now be of the highest st<strong>and</strong>ard. Poor quality or poorly applied<br />
labels will imply a poor beer.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
· The label. Every bottle of beer has at least one label, but frequently several labels are<br />
applied. These can be applied to the body, back <strong>and</strong> neck of the bottle. In addition foil<br />
or plastic capsule tops can be applied over the crown. The function of the label is<br />
twofold: the legislative information required in the country in which the beer is being<br />
sold must be displayed <strong>and</strong> the brewer must display the logo <strong>and</strong> colour detail<br />
associated with the br<strong>and</strong>. The legal data may include a statement of the volume of the<br />
beer in the bottle <strong>and</strong> its strength <strong>and</strong> may in some countries include a statement of the<br />
ingredients. Frequently the shelf-life of the beer must be shown. There are many<br />
different types of paper used for labels; this number has increased with the popularity<br />
of metallized paper labels. There are some basic properties of label paper that can be<br />
described in terms of stiffness, weight, smoothness, density, behaviour in caustic soda<br />
solutions, curl characteristics, etc. (Schwartz, 1997).<br />
Originally label paper was resistant to the caustic soda solution used in the bottle<br />
washer for returnable bottles. This allowed the label to be removed without forming a<br />
pulp that is difficult to dispose of <strong>and</strong>, indeed, in some countries its disposal is<br />
prohibited. In other countries (USA <strong>and</strong> UK) alkali soluble paper is now used <strong>and</strong> pulp<br />
disposal is allowed. Label paper should be resistant to curling <strong>and</strong> creasing <strong>and</strong> often<br />
the reverse side is treated with a pigment to prevent this happening when the front side<br />
is coated to receive the pigments <strong>and</strong> metallized effects. The orientation of the grain of<br />
the paper in relation to the bottle surface is important. The fibres in the bottle label<br />
should run at right-angles to the longitudinal axis of the bottle. If the fibres run parallel<br />
then the labels will tend to come away from the bottles at the edges, a phenomenon<br />
known as `flagging'.<br />
The different types of paper in use can be divided into three categories: paper,<br />
metallized paper, <strong>and</strong> aluminium foil. The paper is usually designated by its weight<br />
per ream, which in the USA contains 480 or 500 sheets. Paper for paper labels is<br />
normally of 40 to 50 lb. (18±23 kg). The surface of an aluminium foil label is 99.5%<br />
pure aluminium at a thickness of 0.009 mm. When applied to paper the overall<br />
thickness is only 0.09 mm, hence very specialized machinery is needed to produce<br />
these labels for application to the bottle. Aluminium foils are often used for neck<br />
labels. The metallized paper label is formed by the vacuum deposition of an ultra-thin<br />
layer of metal on a paper to achieve a `metal' appearance with a much thinner label<br />
than the conventional aluminium foil. There is now a huge choice available to the<br />
brewer <strong>and</strong> this whole area of packaging is subject to continuous development as<br />
brewers strive for br<strong>and</strong> differentiation <strong>and</strong> product enhancement. Cut labels, ready for<br />
use are supplied in storable stacks, which must be kept at a relative humidity of 60 to<br />
70% <strong>and</strong> a temperature of 20 ëC (68 ëF). This prevents curling which otherwise renders<br />
the labels useless for application.<br />
A number of different adhesives have been successfully used to attach labels to beer<br />
bottles (Schwartz, 1997). The most common <strong>and</strong> successful are those based on casein.<br />
These adhesives apply to cold or wet bottles <strong>and</strong> will bond rapidly. They provide good<br />
resistance to condensate water <strong>and</strong> to ice water if bottles are submerged. Easy label<br />
removal in the bottle washer at low caustic strength is achieved. Alternatives have<br />
been based on starch or dextrin. These are cheaper but do not offer such good<br />
properties of, e.g., water resistance. To minimize costs as little adhesive as possible<br />
should be used <strong>and</strong> a rate of 10 g of adhesive/m 2 should be aimed for.<br />
· The labelling machine. Labels are now applied by rotating machines, which can<br />
operate at the speed of fillers. The machine contains a label holder <strong>and</strong> a labeltransporting<br />
device. Modern machines have improved the reproducibility of the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Label<br />
magazine<br />
Gluing pallet<br />
carousel<br />
Data coding<br />
unit<br />
Glue<br />
roller<br />
Gripper cylinder<br />
Label<br />
brush-on<br />
unit<br />
Bottle<br />
carousel<br />
Fig. 21.3 Bottle-labelling station (by courtesy of Krones).<br />
Plan<br />
view<br />
removal of the label from the stack in the holder <strong>and</strong> its application to the bottle.<br />
Labels not squarely applied to the bottle are not tolerated. A glued pallet with a thin<br />
film of glue applied removes the label from the stack in the holder. This results in<br />
simultaneous gluing of the back of the label. The machines now usually employ<br />
oscillating glued pallets with a smooth motion capable of achieving speeds of 70,000<br />
bottles per hour. The overall design relating to the speed of movement of the bottles<br />
<strong>and</strong> the label length is critical. The speed <strong>and</strong> distance between the bottles must be<br />
carefully set when commissioning labelling machines. Careful design <strong>and</strong> commissioning<br />
will allow the use of a label length that is greater than the bottle pitch (distance<br />
between the bottles) with no loss of efficiency.<br />
The labels are normally removed from the pallets <strong>and</strong> applied to the bottles by<br />
gripper cylinders (Fig. 21.3). Body labels <strong>and</strong> neck labels can be transferred in this<br />
way. The body label application occurs when the bottle is smoothly pushed into the<br />
sponge section of the gripper cylinder. If a neck label is to be applied the sponge pads<br />
in the cylinder must be moved out by cams to make contact with the bottle. The<br />
conveying of the bottle through the labeller is critical to ensure that the labels are<br />
correctly aligned. Rotary labellers operate with an infeed star wheel that passes the<br />
bottles to a centring bell which firmly positions the bottle against a plate to receive the<br />
label. To eliminate skewing the bottles are positively clamped between the bottle<br />
plates <strong>and</strong> the centring bells. After application of the body <strong>and</strong> neck labels the bottles<br />
are rotated through 90ë <strong>and</strong> pass a brushing station, which ensures firm contact<br />
between the label <strong>and</strong> the bottle. If a back label is to be applied, the bottle is turned a<br />
further 90ë to meet the back label gripper cylinder <strong>and</strong> the application process is<br />
repeated. After all labels have been applied the bottles are discharged from the labeller<br />
via a star wheel to the discharge conveyor.<br />
Usually a date stamp is now put on the bottle. This might indicate the `best-by' date<br />
or the date of packaging of the beer known as the `born-on' date by one manufacturer.<br />
These dates are sometimes required by law or by the customer but in any event give<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
the drinker aview of the shelf-life of the beer. The date can be put on the label but is<br />
often applied to the bottle itself. The brewer is also likely to include areference mark,<br />
which will be important for product traceability. These reference marks were<br />
traditionally cut into the edge of the label <strong>and</strong> this can be done at manufacture of the<br />
label. Recently other methods have been developed, which include stamping <strong>and</strong><br />
embossing or perforating machines that can operate at up to 60,000 bottles per hour.<br />
These devices have largely been superseded by ink-jet or laser printers. The ink-jet<br />
system uses acharged stream of ink droplets in acontrolled trajectory onto the bottle<br />
surface. The laser directs the energy beam through a metal mask that has the<br />
appropriateinformationcutout.Thebeamisthenfocusedonthelabel,whichsuffersa<br />
change of colour by evaporation to leave the date stamp clearly marked.<br />
Labelling is an important part of the bottling process, vital for the maintenance of<br />
the br<strong>and</strong> image. Developments in label application have now caught up with that of<br />
filling <strong>and</strong> crowning to guarantee the presentation of avery high-class package. This<br />
has been important in re-establishing the bottle against the can as the preferred<br />
package for small pack beer.<br />
21.3.2 Managing beer flow<br />
Managing the beer flow involves bringing the beer to the package in the most efficient<br />
way, consistent with the highest quality being obtained. Almost all bottled beers are now<br />
sterile, filtered products. Sterility is traditionally achieved by pasteurizing the beer in the<br />
bottle after filling <strong>and</strong> crowning. However the beer can be sterilized before filling by<br />
flash pasteurization or sterile filtration. This is what happens with keg beer, which cannot<br />
be tunnel pasteurized because of the huge size of the machine which would be required.<br />
Flash pasteurization is not widely used with bottled beers, although new developments<br />
are occurring (Hyde, 2000), <strong>and</strong> this technique will therefore be considered in the section<br />
on kegging. In managing the beer flow we are concerned with:<br />
· flash pasteurization or sterile filtration<br />
· st<strong>and</strong>ard filling <strong>and</strong> aseptic filling<br />
· crowning<br />
· tunnel pasteurization.<br />
Untilrecentlybeerwasnotmovedlongdistancesforsale<strong>and</strong>somicrobiologicalstability<br />
of the product was not an issue. Modern microbiological stabilization processes began with<br />
the work of Louis Pasteur (Pasteur, 1876), who demonstrated that heating beer to a<br />
sufficientlyhightemperaturewoulddestroybeerspoilagemicrobes.Fortunatelybeerisnota<br />
goodgrowthmedium<strong>and</strong>onlysupportsslowgrowthofarelativelysmallnumberofmicrobes<br />
(Rainbow,1971),whichdoesnotincludepathogens(Bunker,1955;Chapter17)).Thetaskof<br />
providing the customer with amicrobiologically stable product is therefore simplified.<br />
Absolute security in the sterility of bottled products is required. Bottles may be<br />
distributed over very long distances, <strong>and</strong> often have shelf-lives of 40 to 52 weeks. The<br />
traditional <strong>practice</strong> has been to pasteurize the beer after filling in its final package (see<br />
later). However this process is very energy intensive, requiring at least 1,000 MJ/1,000<br />
bottles. Beer flavour is adversely affected by pasteurization <strong>and</strong> probably suffers even if<br />
oxygen contents are kept to levels < 0.1 mg/l. This has led to the use of filtration to<br />
sterilize the beer prior to filling. Sterile filtration physically removes organisms from the<br />
beer. The technique dem<strong>and</strong>s subsequent aseptic filling <strong>and</strong> the application of rigorous<br />
st<strong>and</strong>ards to ensure no contamination enters the bottle.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Sterile filtration<br />
Some ofthe techniquesused toachieve non-biologicalstability (Chapter 15)ofbeerhave<br />
been used <strong>and</strong> claimed to yield sterile beer. So plate <strong>and</strong> frame filters using pulps of<br />
kieselguhr or perlite or sheet filters using asbestos/cellulose sheets are said to be effective<br />
for sterile filtration as well (Wilson, 1997). This type of sterile filtration was originally<br />
described in the 1930s by the Seitz Company that developed the EK sheet (Entkeimung ˆ<br />
sterilization). These sheets were made of a mixture of cellulose <strong>and</strong> asbestos <strong>and</strong> were<br />
4.5 mm thick with a pore size of 5±20 m. Flow rate achievable was 1.5 hl/m 2 /h at a<br />
maximum pressure of 1.5 bar. The quality <strong>and</strong> composition of the sheets has changed<br />
much in recent years (Chapter 15) <strong>and</strong> beer stabilizers can now be incorporated into the<br />
sheet such as PVPP (polyvinylpolypyrrolidone) or silica hydrogel.<br />
These pulp <strong>and</strong> sheet methods act by depth filtration. The pad or sheet contains<br />
millions of flow channels. Micro-organisms are mechanically entrapped or absorbed as<br />
the beer flows through the filter. There is also an adsorptive effect as microbes are fixed<br />
by electric charge. The organisms bear a negative charge, which attracts to the positive<br />
charge of the filter matrix. These methods have been used for sterile filtration of beer.<br />
However, since the carcinogenic properties of asbestos were recognized, asbestos was<br />
replaced in sheet filters with kieselguhr or perlite. Some brewers lost confidence in depth<br />
filtration as a method to guarantee beer sterility. The situation was changed with the<br />
availability of membrane filters.<br />
Membrane filters can be classified as surface filters which operate on a sieving method<br />
for the removal of organisms (Wilson, 1997). The membrane is a uniform continuous<br />
structure with regularly spaced uniformly sized pores. The membranes are made of<br />
cellulose esters <strong>and</strong> are normally 150 m thick (Bush, 1964). As beer passes through the<br />
filter all organisms larger than the pore size of the filter are trapped <strong>and</strong> retained on its<br />
surface. For brewery use a pore size of 0.45 m is necessary to retain all potential<br />
spoilage microbes. Membrane filtration is the only sterile filtration method that will<br />
provide absolute sterility. However, membranes are prone to blockage <strong>and</strong> it is essential<br />
that the beer presented to the membrane has received satisfactory primary filtration. The<br />
beer must be free of particles that will blind the sterilizing filter, the sole aim of which is<br />
to achieve sterility. This is now achieved by a sequence of filters after the primary<br />
kieselguhr filtration. These are frequently cartridge filters. There can be two or three in<br />
line with reducing pore sizes, e.g.,<br />
· kieselguhr filtration<br />
· cartridge filter 1, 5 m pore size<br />
· cartridge filter 2, 1 m pore size<br />
· sterilizing membrane filter, 0.45 m pore size.<br />
Using a system of this type, high throughput <strong>and</strong> sterility can be achieved. A nondestructive<br />
test to indicate the suitability of beer for membrane filtration has been<br />
described (Pall, 1975). It is also important to have sound microbiological control<br />
throughout the brewery so that the effectiveness of the sterile filter is further enhanced.<br />
Other types of filter have been described (Moll, 1994). Ceramic c<strong>and</strong>les have been<br />
used in Japan. Flow rates of 10 hl/m 2 /h were achieved with ceramic c<strong>and</strong>les of 25 mm<br />
wall thickness <strong>and</strong> a pore size of 25 m (Beer, 1989). Cross-flow filtration methods have<br />
also been tested, where the liquid flow is tangential to the filter medium (Atkinson,<br />
1988). Tubular membranes <strong>and</strong> s<strong>and</strong> have been used as the filter medium. These systems<br />
are not fully developed for beer <strong>and</strong> have been prone to blockage with nonmicrobiological<br />
polymers present in the beer such as -glucans. If the pore size is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
educed to below 0.45 m, which has been the case in some systems, bittering substances<br />
are removed from the beer (Donhauser <strong>and</strong> Jacob, 1988). Absolute sterility has not been<br />
consistently achieved on the industrial scale using cross-flow methods.<br />
Sterile filtration has the advantage over pasteurization of giving very gentle treatment<br />
to the beer; with no heating <strong>and</strong> cooling there is no potential for flavour changes. The<br />
additional filtration can also improve non-biological stability <strong>and</strong> clarity of the beer. The<br />
technique has lower capital <strong>and</strong> operating costs than tunnel pasteurization. Sterile<br />
filtration must be operated with sterile filling <strong>and</strong> so this obviously adds to the difficulty<br />
of the filling operation. There must be abuffer tank between the sterilizing filter <strong>and</strong> the<br />
filler <strong>and</strong> aconstant flow must be obtained to avoid pressure shocks. The pressure drop<br />
across the filter should be monitored. A drop in pressure of 2 bar (30lb./in. 2 ) is<br />
acceptable. Care must be exercised when changing tanks to be filtered or when filtering<br />
tanks at low beer levels. Gas bubbles can break out at the filter pump which can give a<br />
sudden pressure drop <strong>and</strong> hence aloss of sterility in the system. This can be avoided with<br />
proper attention to the monitoring of pressures <strong>and</strong> flows.<br />
Sterile filtration must be operated in asanitary system. Effective CIP is essential. The<br />
system must be cleaned, sanitized, <strong>and</strong> back flushed using hot <strong>and</strong> cold de-aerated water.<br />
Normally water at 80ëC (175ëF) is circulated for 60 minutes. The CIP system is an<br />
integral part of the whole sterile filter plant.<br />
One source of competitive advantage to the brewer is the delivery of the freshest<br />
tasting beer possible. This is most likely to be achieved using sterile filtration <strong>and</strong> filling.<br />
Consequently there is great interest in perfecting these techniques.<br />
St<strong>and</strong>ard filling<br />
The filler is the most important piece of equipment in the bottling line. For optimum<br />
efficiencythefillermustruncontinuouslythroughouttheshift.Machinesbefore<strong>and</strong>after<br />
thefillermustbedesignedtosupplybottles<strong>and</strong>takethemawayatleastasfastastheyare<br />
filled <strong>and</strong> sealed. The beer supply system must keep carbon dioxide in solution <strong>and</strong><br />
exclude oxygen. The whole filling operation must not add more than 0.03mg/l dissolved<br />
oxygen to the beer. To maintain the carbon dioxide content achieved in the bright beer<br />
tank the pressure must be maintained during filling at one bar above the CO2 saturation<br />
pressure. Contamination of the beer must be avoided. Microbiological sterility will be<br />
achieved by tunnel pasteurization or by aseptic filling if the beer has been sterile filtered<br />
or flash pasteurized.<br />
The objectives in filling bottles are to preserve the quality of the beer <strong>and</strong> get the right<br />
volume of liquid into the bottle. As beer contains carbon dioxide, beer fillers always<br />
operate at ahigh pressure <strong>and</strong> are generally called counter-pressure fillers. To deliver the<br />
required volume of beer, beer fillers are designed either to fill to alevel in the bottle or to<br />
fill by displacement. Bottle fillers are rotary machines <strong>and</strong> can rotate clockwise or anticlockwise.<br />
Machines can now contain up to 200 filling valves <strong>and</strong> can fill at 1,000 to<br />
1,600bottlesper minute.Most fillers perform inasimilarbasicway (Fig.21.4, plan view<br />
<strong>and</strong> Fig. 21.6, sequence of operation):<br />
· beer is received<br />
· bottles are positioned in predetermined spacing under the filler heads on the filler<br />
platform<br />
· bottles are lifted up to the filler head<br />
· bottles are evacuated <strong>and</strong> counter-pressured with carbon dioxide<br />
· bottles are filled<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
· corrections are made to fill height<br />
· pressure is released<br />
· bottles are lowered <strong>and</strong> moved onto the crowner, where they are sealed.<br />
Therefore in sequence we have:<br />
7<br />
6<br />
8<br />
Fig. 21.4 Bottle filling, plan view: 1, 1st evacuation; 2, carbon dioxide flushing; 3, 2nd evacuation;<br />
4, pressurizing; 5, filling; 6, filling completed <strong>and</strong> settling; 7, correction; 8, snifting<br />
(by courtesy of Krones).<br />
· Beer supply. Beers are almost always cold filled into bottles at 0to 5ëC (32±40ëF).<br />
To maintain the carbon dioxide in solution the beer should be at one bar pressure (15<br />
lb./in. 2 )in the supply to the filler bowl. This is normally achieved by maintaining a<br />
carbon dioxide top pressure in the headspace of the bright beer tank. There must be no<br />
dropinpressureatthetankoutlet<strong>and</strong>sothebeerflowrateshouldnotexceed 225cm/s<br />
uptothispoint.Thebeershouldnothavetotraverseright-angle bends(Spargo,1997).<br />
The volume of beer supplied to the fillercan be measured in line with amagnetic flow<br />
meter provided there is enough straight pipe-work in the flow line for the beer to<br />
achieve laminar flow. All fillers have a`bowl' or tank to receive beer from the bright<br />
beer tank <strong>and</strong> to distribute the beer to the filler heads. On large machines (>500<br />
bottles/min.) the bowl is annular <strong>and</strong> the beer in the bowl is counter-pressured from<br />
abovewithcarbondioxide.Asensor,whichisusuallyafloatvalve,controlstheheight<br />
ofthebeerinthebowl.Fillingvalvesareusuallymountedatthebottom ofthebowlor<br />
attached to the outside of the bowl. In this way beer distribution pipe-work to the<br />
filling valve can be eliminated.<br />
· Bottle positioning <strong>and</strong> lifting (Fig. 21.5). The lifting platforms place the bottles<br />
against the filling heads to provide asecure connection. This is done by compressed<br />
air. The same device using aroller lowers bottles. Compressed air can be forced back<br />
in the lowering process into the supply pipe <strong>and</strong> so is not lost. The device to position<br />
the bottle is called acentring bell or tulip. This is carried up by the bottle being raised.<br />
The centring tulip ensures that the bottle is centred in the filling head <strong>and</strong> sealed.<br />
· Bottle evacuation, counter-pressure <strong>and</strong> filling (Fig. 21.6). There are anumber of<br />
different beer filling heads. Beer can be introduced into the bottle via afilling tube.<br />
The filling tube reaches to the bottom of the bottle. Obviously the tube must have a<br />
smaller diameter than the bottle opening <strong>and</strong> this limits filling speed. Filling tubes<br />
5<br />
1 2<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
3<br />
4
c<br />
b<br />
a<br />
d<br />
Carbon dioxide<br />
Enlargement<br />
Fig. 21.5 Bottle filling, section view; basic position before the bottle is lifted onto the filling<br />
head (air in bottle); a, vent tube; b, siphon type gas lock; c, correction channel; d, control cylinder<br />
for liquid valve <strong>and</strong> pressurizing; e, gas needle; f, vacuum/CIP channel; g, snift channel<br />
(by courtesy of Krones).<br />
introduce beer just above the bottom of the bottle. There is very little uptake of<br />
oxygen. The filling tube is made up of atube through which the beer passes <strong>and</strong> a<br />
return gas vent at the level of the required filling height. Some high-speed fillers<br />
(>1,000 bottles/min.) operate without filling tubes. The beer is introduced down the<br />
side of the bottle neck. There is adanger of ingress of oxygen with this system but the<br />
system will operate at higher speed than the filling tube system. Asequence of<br />
operation for anon-filling tube machine is shown (Fig. 21.6). After lifting the bottle is<br />
evacuated (21.6a) <strong>and</strong> about 90% vacuum is achieved. The next phase is aflushing<br />
with carbon dioxide (21.6b), which enters the bottle from the filling bowl. There is<br />
then normally asecond evacuation (21.6c). The vacuum allows the displacement of<br />
the carbon dioxide <strong>and</strong> some air, which will still be present. The secondary evacuation<br />
stage would not be carried out with afilling tube bottle filler. Counter-pressurization<br />
with carbon dioxide then takes place (21.6d). The process is similar to the carbon<br />
dioxide flush (21.6b) but takes longer <strong>and</strong> ahigh concentration of the gas is achieved<br />
in the bottle. The pressure in the bottle is now the same as that in the filler bowl. Beer<br />
can now flow downwards against the bottle sides in athin film displacing the carbon<br />
dioxide(21.6e).Assoonasthebeerreachestheendoftheventtubeitrisesinthistube<br />
(21.6f) <strong>and</strong> the gas above cannot escape. There is an overfill in this stage which must<br />
be corrected (21.6g). The beer valve is closed <strong>and</strong> the gas valve remains open, carbon<br />
e<br />
Beer<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
f<br />
g
1st evacuation<br />
Vacuum in bottle<br />
Carbon dioxide<br />
Beer<br />
Fig. 21.6 (a) 1st evacuation of bottle (vacuum in bottle) (by courtesy of Krones).<br />
CO 2 flushing<br />
Carbon dioxide<br />
Carbon dioxide<br />
Fig. 21.6 (b) Carbon dioxide flushing (carbon dioxide in bottle) (by courtesy of Krones).<br />
Beer<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
2nd evacuation<br />
Vacuum in bottle<br />
Carbon dioxide<br />
Beer<br />
Fig. 21.6 (c) 2nd evacuation (vacuum in bottle) (by courtesy of Krones).<br />
Pressurization<br />
Carbon dioxide<br />
Carbon dioxide<br />
Beer<br />
Fig. 21.6 (d) Pressurization (carbon dioxide in bottle) (by courtesy of Krones).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Filling<br />
Beer displaces<br />
carbon dioxide<br />
from bottle<br />
Carbon dioxide<br />
Fig. 21.6 (e) Filling (beer displaces carbon dioxide from bottle) (by courtesy of Krones).<br />
Filling completed<br />
Beer<br />
Beer<br />
Carbon dioxide<br />
Beer<br />
Fig. 21.6 (f) Filling completed (beer in bottle) (by courtesy of Krones).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Correction<br />
Beer<br />
Carbon dioxide<br />
Fig. 21.6 (g) Correction of bottle contents (excess beer displaced back to filling bowl) (by<br />
courtesy of Krones).<br />
Snifting<br />
Beer<br />
Beer<br />
Carbon dioxide<br />
Beer<br />
Fig. 21.6 (h) Snifting (controlled pressure release) (by courtesy of Krones).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
dioxide at asmall overpressure enters the neck of the bottle <strong>and</strong> the beer above the<br />
ventpipe isforced back intothefiller bowl.Anexact fillingheightcan sobeobtained.<br />
ThereisfinallyacontrolledpressurereleaseknownintheUKasa`snift'(21.6h).This<br />
takes place before breaking the seal with the valve <strong>and</strong> the centring tulip. Avent valve<br />
is opened <strong>and</strong> excess pressure is released so atmospheric pressure is achieved <strong>and</strong><br />
foaming of the contents is avoided. The gas movements in stage hwere formerly<br />
controlled using cams, which were very precisely cut; cams have now been replaced<br />
by electronic systems. After filling the bottle is lowered from the filling head <strong>and</strong><br />
conveyed to the crowner.<br />
It is essential en route to the crowner to eliminate air from the head space of the bottle<br />
to avoid subsequent oxidation of the beer. This is now usually done by water-jetting. A<br />
high-pressurestreamofsterilizedwaterissprayedontoeachopenbottle.Onlyafew lof<br />
water enter the bottle but this causes an effective beer foaming, which rises in the neck<br />
<strong>and</strong> dispels oxygen <strong>and</strong> prevents any further entry. This process is carefully adjusted to<br />
minimize beer loss. Liquid nitrogen jetting may also be used (Donovan et al., 1999). This<br />
technique reduces beer losses <strong>and</strong>, perhaps more importantly, reduces waste. The foam is<br />
formed initially around the nitrogen gas <strong>and</strong> it is not necessary to expel foam containing<br />
air <strong>and</strong> losses are lowered.<br />
Aseptic filling<br />
Sterile filtration is apowerful technique, which can provide beer in its freshest form.<br />
However the technique must be combined with aseptic filling. Aseptic filling requires<br />
considerable expertise to achieve success. Anumber of fundamental requirements must<br />
be met. All employees involved in sterile filling must be well trained <strong>and</strong> be determined<br />
to make it work. The beer to be packaged must be free of organisms <strong>and</strong> therefore must<br />
be presented from awell managed sterile filtration plant or from flash pasteurization. If<br />
sterile filtered this has probably involved afinal pass through a0.45 membrane filter.<br />
The line must be mechanically reliable. Generally the most satisfactory operation is<br />
achieved from long runs. If there are frequent breakdowns then re-sanitization is required<br />
with consequent risk to the product <strong>and</strong> poor efficiency. There must be clearly defined<br />
<strong>and</strong> written-down sanitation procedures. These must include the philosophy on the<br />
method of sanitation be it hot water, steam, or chemical.<br />
There must be an effective microbiological control system in place (see also Chapter<br />
17). This is best achieved by the operators on the line being properly trained in<br />
microbiological sampling <strong>and</strong> analysis. In this way the operator comes to `own' the<br />
problem of the line <strong>and</strong> problems are more likely to be solved. The main focus of the<br />
sampling plan is normally the packaged beer. Two samples of beer should be taken off<br />
each filler head for every two hours of filling. These can be analysed for contamination<br />
by forcing <strong>and</strong> membrane filtration. Attention should also be paid to the plant, the CIP<br />
system <strong>and</strong> the water in use.<br />
These basic principles <strong>and</strong> operating philosophy of the equipment are more important<br />
in achieving success with sterile filling than the detailed design of the plant. The design<br />
must be kept simple <strong>and</strong> open. The main objective in pipework is to avoid sharp bends<br />
likely to restrict the flow of cleaning materials. Pipework <strong>and</strong> tanks should be made of<br />
stainless steel <strong>and</strong> should be capable of withst<strong>and</strong>ing a flow of hot water at 80 ëC (175 ëF)<br />
for 30 minutes. Steam can be used for re-circulatory cleansing but generally is not as<br />
effective as hot water as it lacks an attrition effect. The filler must be sanitized within one<br />
hour of bottling start-up. The best systems have a dedicated CIP system. The water<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
should flow in the reverse direction to the beer. A final chilled water rinse is usually used<br />
<strong>and</strong> to this water can be added a 25 ppm iodophor solution. This can be left in the system<br />
<strong>and</strong> drained immediately prior to the start of a filling run.<br />
There has been discussion of the desirability of enclosing sterile filling plant in a<br />
sterile room. Some systems work successfully in the open in a packaging hall. However,<br />
there are advantages in enclosure. One is to show operators <strong>and</strong> visitors, including<br />
customers, that something special is taking place. A carefully designed enclosure allows<br />
the control of a sterile air supply in both temperature <strong>and</strong> humidity. It also facilitates the<br />
external cleaning of the plant. Sterile filling runs are usually successful over periods of 12<br />
to 36 h. During this time a continuous sanitation programme is necessary. General<br />
housekeeping must include the continuous removal of broken bottles <strong>and</strong> the spraying of<br />
the floor to wash away spilt beer. Foam cleaning the plant with iodophor solution during<br />
operation is beneficial to hygiene. Close attention should be given to the filler tubes <strong>and</strong><br />
the crowner platforms (see later). If production stops during a bottling run then iodophor<br />
treatment should continue. If the stoppage exceeds one hour then the plant should be shut<br />
down <strong>and</strong> a full sanitation carried out. The filler must be cleaned at the end of the run by a<br />
hot water flush prior to full sanitation one hour before the next run.<br />
The capital cost of a sterile filtration/sterile filling plant can be 10% higher than the<br />
costs for a tunnel pasteurizer. However, operating costs are much lower <strong>and</strong> when<br />
assessing all the costs, sterile filling is about half the cost of tunnel pasteurization (Hyde,<br />
2000). Sterile filtration avoids the potential flavour deterioration effects of pasteurization<br />
<strong>and</strong> gives the consumer the freshest tasting beer. This can create potential competitive<br />
advantage. However, training <strong>and</strong> education of operators <strong>and</strong> acute attention to detail is<br />
essential for success. In addition the throughput of the line is not likely to be as good as<br />
with normal filling because of the need to sanitize after stoppages.<br />
Crowning<br />
The bottles must be quickly closed after filling. The closure machine is therefore an<br />
integral part of the filling machine <strong>and</strong> comes in the line immediately after the point when<br />
air has been dispelled from the bottle neck by water-jetting. Most beer bottles are closed<br />
with crown corks. The crown cork evolved from the natural cork stopper. Cork stoppers<br />
were effective closures but did not meet the need to be applied at speed. The automatic<br />
production of glass bottles was perfected in the USA in 1903 (Everett, 1997). This<br />
provided bottles with uniform dimensions at the neck that could take a st<strong>and</strong>ard closure.<br />
William Painter had already filed US Patent 468226 in 1892 for the `crown closure'.<br />
Painter's crown was lined with cork <strong>and</strong> characteristically had corrugations in the `skirt'.<br />
Subsequently a crown with no corrugations was designed along with a machine which<br />
applied the closure to the bottle <strong>and</strong> formed the familiar corrugations.<br />
The cork was mostly obtained from oak trees in the western Mediterranean <strong>and</strong> during<br />
the Second World War supplies ceased <strong>and</strong> prices rose. At this time plastic alternatives to<br />
the cork lining were developed. The first PVC (polyvinylchloride) lining was introduced<br />
in 1955. Nowadays, the crown cork is made of tinned or chromed steel or stainless steel<br />
plate 0.235 mm thick. The brewery logo is usually displayed on the outer surface. The<br />
st<strong>and</strong>ard crown cork has 21 corrugations <strong>and</strong> an outer diameter of 32.1 mm, an internal<br />
diameter of 26.75 mm <strong>and</strong> a height of 6.0 mm. The normal lining is plasticized PVC. A<br />
lubricant is included in the lining for `twist-off' crowns so they can be removed by h<strong>and</strong>.<br />
The lining does not affect flavour <strong>and</strong> is approved for use in most countries. The crown<br />
must retain the pressure in the bottle until opened <strong>and</strong> there must be no gaseous exchange<br />
with the atmosphere.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 21.7 Turning tube for orientating crown corks for bottle closure, incorrectly positioned<br />
corks are turned up to 180ë to gain the correct orientation for bottle closure (Kunze, 1999).<br />
Crown corks from stock are conveyed to a storage hopper on the closure machine <strong>and</strong><br />
a sorting device brings them into the correct orientation for closure onto the bottle.<br />
Incorrectly orientated crowns are re-positioned in a turning tube (Fig. 21.7) containing<br />
fixed inserts before falling downwards through a groove. This device is simple <strong>and</strong><br />
effective but is subject to heavy wear <strong>and</strong> so it must be replaced regularly to avoid<br />
frequent stoppages as crowns become jammed in the tube. Crowns are delivered to the<br />
closure device either mechanically or pneumatically, preferably using carbon dioxide or,<br />
less desirably, air. A magnet in the ejector then holds the crown cork. The bottles are<br />
delivered to the crowner by a star-wheel <strong>and</strong> the crowner head, containing the correctly<br />
orientated crown, descends onto the bottle. The crown then makes contact with the crown<br />
ring of the bottle. A spring forces closure onto the bottle <strong>and</strong> the 21 corrugations of the<br />
crown cork are bent down onto the bottle neck. A gas tight seal is obtained. The bottles<br />
are usually sprayed with water to blow off any beer residues arising from the removal of<br />
head-space air.<br />
Inevitably, some returnable bottles are damaged <strong>and</strong> this damage is sometimes to the<br />
neck of the bottle. This can affect the security of the crown cork closure. In one case this<br />
resulted in as many as 800 defective closures per 1 million bottles by one Company<br />
(Duffy <strong>and</strong> du Toit, 2000). Development work on the crowning machine followed by<br />
improved preventative maintenance led to a reduction in defective closures to five to<br />
eight per million <strong>and</strong> a consequent enhancement of the presentation of the br<strong>and</strong> whilst<br />
retaining the environmental friendliness of the returnable bottle.<br />
Sometimes other bottle closures are used. A `rip-off' closure has been developed,<br />
made from light gauge aluminium that can be applied in modified crown cork closure<br />
machines. This type of cap has also been used on wide-mouth bottles that are easier to<br />
drink from. Roll-on closures have also been used on 32 oz. bottles in the USA. These are<br />
normally pilfer proof where the cap is held to a ring by uncut metal `bridges' which<br />
fracture when the cap is turned by h<strong>and</strong> to open the bottle. The ring remains on the bottle.<br />
Largely because of waste disposal reasons these closures have not been popular in Europe<br />
except on two-litre PET (see later) bottles.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Tunnel pasteurization<br />
The surest way to provide the customer with beer containing no viable micro-organisms<br />
is to treat it at the last possible moment. That is, the beer is treated in its package after<br />
closure. The beer is not exposed to the atmosphere until it is consumed. This is achieved<br />
by pasteurization (Pasteur, 1876). Pasteurization is the killing of micro-organisms in<br />
aqueous solutions by heat. Beer can be pasteurized in bulk while flowing, which is known<br />
as flash pasteurization, or in the package, which is known as tunnel pasteurization. Flash<br />
pasteurization occurs before the beer is put into its final container <strong>and</strong> so is an alternative<br />
to sterile filtration. Flash pasteurization is usually associated with the preparation of keg<br />
beer. Tunnel pasteurization is most commonly used with small-pack beer, either bottles<br />
or cans. The theory of pasteurization is common to both systems.<br />
· Theory of pasteurization. The basis of pasteurization is establishing the minimum<br />
time <strong>and</strong> temperature required to destroy all expected biological contaminants at the<br />
highest concentrations at which they may occur in filtered beer. Different food products<br />
have different requirements for pasteurization, <strong>and</strong> those that can contain sporeforming<br />
bacteria require much higher heat treatment than beer. Mixed populations of<br />
common brewery contaminating organisms were subject to a range of times <strong>and</strong><br />
temperatures in beer (Fig. 21.8, known as a lethal rate curve) <strong>and</strong> were examined for<br />
subsequent viability. Typically at temperatures of over 50 ëC (122 ëF) an increase in<br />
temperature of 7 ëC (12.5 ëF) accelerated the rate of cell kill by ten times. Therefore:<br />
· 53 ëC: minimum time to kill population 56 min.<br />
· 60 ëC: minimum time to kill population 5.6 min.<br />
· 67 ëC: minimum time to kill population 0.56 min.<br />
One pasteurization unit (PU) for beer has been arbitrarily defined as the biological<br />
destruction obtained by holding a beer for one minute at 60 ëC (140 ëF) (Del Vecchio<br />
et al., 1951). Therefore in Fig. 21.8 the point at which the line crosses the 60 ëC line<br />
Fig. 21.8 The effect of time <strong>and</strong> temperature on the viability of a mixed population of yeasts<br />
<strong>and</strong> brewery bacteria. The hatched area shows the range of conditions where all cells are killed<br />
(Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
gives the thermal resistance of the particular suspension of organisms, this is 5.6 min<br />
<strong>and</strong> so to achieve effective pasteurization the holding time at 60ëC (140ëF) must<br />
exceed 5.6min. The slope of the line in Fig. 21.8 is known as the Zvalue. The lethal<br />
effect (PU) is simply the product of the lethal rate <strong>and</strong> the time of application. The<br />
lethal effect at various temperatures in aprocess is additive, therefore the sum of the<br />
lethal effect is the quantity of sterilization achieved:<br />
Lethal Effect ˆLxt…PU†<br />
where LˆLethal Rate <strong>and</strong> tˆtime held at temperature TëC <strong>and</strong><br />
Lˆ1=Log 1 …60 T/Z†<br />
Under laboratory conditions beer can be sterilized by treatment with 5±6 PUs when<br />
cell numbers are 0.2mg/l. The most heat resistant organisms are the lactic acid<br />
bacteria <strong>and</strong> some Saccharomyces species such as S. pastorianus. The pasteurizationofreturnedbeer<br />
ismoredifficultbecause thiscan containveryhigh numbersof<br />
organisms. It is usually only blended into fresh beer at low rates (50PU)<br />
are minimized.<br />
· Practice of tunnel pasteurization. Atunnel pasteurizer comprises alarge metal-cased<br />
enclosure through which the bottles are passed by aconveying system (Fig. 21.9). The<br />
conveying system can use aslat continuous conveyor chain on which the bottles are<br />
slowly moved or can use awalking beam conveyor. In awalking beam the bottles<br />
st<strong>and</strong> on bars <strong>and</strong> are slowly moved forward in cyclical steps (Fig. 21.10). The<br />
walking beam has the advantage that the moving elements remain in the same<br />
temperature zone whereas the continuous conveyor inevitably carries some heat <strong>and</strong><br />
water away to another zone. The pasteurizer operates as a series of zones. The bottles<br />
are loaded at one end <strong>and</strong> then conveyed under a series of water sprays. The sprays are<br />
arranged such that the bottles are subject to increasingly hot water until the beer in the<br />
bottles reaches the pasteurization temperature. In tunnel pasteurizers this temperature<br />
is usually 60 ëC (140 ëF), which is held for 20 min. This delivers 20 PU to the beer. The<br />
bottles then move to a cooling zone where they are subjected to cooling by cold water<br />
sprays. The bottles then leave the pasteurizer.<br />
Most modern tunnel pasteurizers are double-deck machines (Fig. 21.9). The bottles<br />
travel to the end of the top deck of the pasteurizer <strong>and</strong> then go down to the lower deck.<br />
In this system the water in the sprays on the top deck pre-heats the bottles <strong>and</strong> passes<br />
to the lower deck to be used for cooling. Exact control of temperature is important.<br />
The objective is the effective kill of all organisms with the minimum use of energy.<br />
Temperatures in all the zones are recorded <strong>and</strong> often computed with the time to show<br />
pasteurization units supplied. Passage through the machine normally takes one hour.<br />
Tunnel pasteurizers are the biggest piece of equipment in the packaging line <strong>and</strong> a<br />
surface area of 3.5 m 2 /1000 bottles/h is needed to ensure effective pasteurization.<br />
Modern machines can achieve outputs of at least 150,000 bottles/h. The water used in<br />
the pasteurizer must be clean <strong>and</strong> kept at a pH value of around 8. If not the pasteurizer<br />
will become grossly infected with a variety of bacteria <strong>and</strong> moulds <strong>and</strong> the sprays will<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Load end<br />
1st<br />
pre-heat 2nd pre-heat Pasteurizing zone Pre-cool Cooling<br />
Pre-cool<br />
reservoir<br />
2nd pre-heat<br />
reservoir Pasteurizing reservoir<br />
1st<br />
pre-heat<br />
reservoir<br />
Overflows<br />
Cooling<br />
reservoir<br />
Discharge end<br />
Service<br />
water<br />
Fig. 21.9 General arrangement of a double-deck tunnel pasteurizer. Times <strong>and</strong> temperatures for the various zones: 1st pre-heat, 5 min. at<br />
35±50 ëC; 2nd pre-heat, 13 min. at 50±62 ëC; Pasteurization, 20 min. at 60 ëC; Pre-cool, 5 min. at 60±49 ëC; Cool, at 49±30 ëC; Discharge,<br />
2 min. at 30±20 ëC. For can pasteurization the pre-heat <strong>and</strong> cooling zones may be shortened in length <strong>and</strong> therefore the beer spends less<br />
time in these zones, because of the lower structural strength of the can (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Fig. 21.10 Walking beam principle for moving bottles through a tunnel pasteurizer. Left-h<strong>and</strong><br />
sequence represents side elevations showing successive movements of shaded grid beams in relation<br />
to unshaded grid. Right-h<strong>and</strong> sequence represents corresponding plan views (D indicates<br />
downward movement <strong>and</strong> U upward movement), (Coleman, 1976 <strong>and</strong> Hough et al.,1982).<br />
be blocked <strong>and</strong> the bottles will emerge dirty. Water hardness must be removed or at<br />
the very least calcium ions must be sequestered to keep them in solution so that the<br />
bottles do not dry with a coating of calcium salts. Sometimes bacteriocides containing<br />
chlorine are added to the water.<br />
During pasteurization pressure builds up in the headspaces of the bottles. The exact<br />
pressure build-up depends on the volume of the headspace, which must be accurately<br />
controlled by fill height to limit possible bottle breakage. Carbon dioxide can also be<br />
released from its supersaturated state in the bottle if the bottles are knocked during<br />
pasteurization. If the bottle is defective the bottle may burst or carbon dioxide may<br />
leak allowing a reduction in the gas content of the beer. For example, if the carbon<br />
dioxide content of the beer is 0.38% <strong>and</strong> the percentage headspace in the bottle is<br />
1.7%, the pressure would exceed 10 bar during pasteurization if supersaturation did<br />
not hold. A walking beam conveyor is generally better at avoiding bottle knocks than a<br />
slat conveyor. If effective controls of fill height are in place bottle breakage should not<br />
exceed 0.2% of the total bottles processed.<br />
Tunnel pasteurization is a very safe method of assuring sterile beer. It is, however,<br />
the most expensive method of assuring sterility in both capital <strong>and</strong> operating costs<br />
(Hyde, 2000), which are double those of sterile filtration <strong>and</strong> at least five times more<br />
than for flash pasteurization (see later). It is also a fault of tunnel pasteurizers that it is<br />
easy to overpasteurize the beer. The beer will then not taste fresh <strong>and</strong>, particularly if<br />
the dissolved oxygen level was > 0.2 mg/l, will very quickly develop a cardboard<br />
flavour. This has stimulated interest in sterile filtration <strong>and</strong> flash pasteurization for<br />
small pack beer.<br />
21.3.3 Managing plant cleaning<br />
All areas of a packaging plant must be kept clean, using a rigorous housekeeping regime.<br />
This is particularly important in a bottling plant, because breakages of bottles will occur<br />
<strong>and</strong> the broken glass is a threat to employees <strong>and</strong> the product. It must be removed at once.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
U<br />
U<br />
U<br />
U<br />
U<br />
U<br />
D<br />
D<br />
D<br />
D<br />
D<br />
D
The floor area around the filler should be disinfected daily. Thorough cleaning of the<br />
filler <strong>and</strong> crowner is essential. Infection can develop in the filler; usually as aresult of<br />
Acetobacter infection, which develops on residues of beer <strong>and</strong> foam. Slime caps are<br />
formed which can provide encouragement to other organisms particularly Lactobacillus<br />
sp., which will grow in beer at very low oxygen levels. This may result in aneed for<br />
increased pasteurization with aconsequent adverse effect on beer flavour.<br />
Thorough mechanical cleaning of the filler <strong>and</strong> the provision of ahot water flush unit<br />
avoids infection. After bottles have left the filler this device can provide an overflow of<br />
hot water at 90ëC (195ëF) in the filling machine, at the point of delivery of the crowns<br />
<strong>and</strong> over the star wheels. The sequence can be arranged so this overflow occurs every 2±<br />
3hours, during 2±3 filler rotations at half speed. This should be supported with foam<br />
cleaning of the conveyors <strong>and</strong> spraying with iodophor at the end of each bottling run.<br />
The key to successful plant hygiene is management <strong>and</strong> training. Even if tunnel<br />
pasteurization is to take place after filling this is not an excuse for poor hygiene<br />
management.<br />
21.3.4 Materials for making bottles<br />
The most important material for both returnable <strong>and</strong> non-returnable beer bottles is glass.<br />
Glass exists at room temperature as asuper-cooled liquid. It is chemically inert <strong>and</strong> will<br />
not add or take away properties from the product it contains. Glass will only break under<br />
tensile load <strong>and</strong> the compressive strength of glass fibres is such that they will withst<strong>and</strong><br />
34,000 bar. Glass is recyclable <strong>and</strong> cullet, broken glass fragments, is an important<br />
constituent of new-made glass (Moll, 1997).<br />
Glass bottles can be made in several different colours but the choice of colour is<br />
importantforthe flavourstability of the beer. White lightcan act onthe iso -acids inbeer<br />
to form acompound, 3-methyl-2-butene-1-thiol. This imparts an aroma <strong>and</strong> flavour known<br />
as light-struck or `skunky', <strong>and</strong> renders the beer very unpleasant. It can be detected by<br />
some people at aconcentration of 0.4 parts per trillion (0.4ng/l)! Beer must therefore be<br />
protected from light. Brown glass is the best option <strong>and</strong> is much better than green glass in<br />
this respect. The worst type of glass is clear (flint) glass in which the flavour of beer<br />
deteriorates very rapidly when exposed to light. Marketing departments frequently want to<br />
use green or flint glass for beer considering that it gives the beer amore attractive<br />
appearance. This is resisted by brewers concerned about flavour risks. This is one reason<br />
why reduced isohumulones are now added to some beers. These compounds do not break<br />
down under light to the thiol. The compounds also have increased hydrophobicity<br />
compared to isohumulone <strong>and</strong> so they also enhance foam (Chapter 19).<br />
Bottles can be made to very exacting specifications, which are often controlled by<br />
statute. Bottles can be assessed by a number of quality control methods to ensure<br />
satisfactory performance on the line (Moll, 1997). Consumers prefer glass for the<br />
packaging of small pack beer believing it gives a more upmarket image <strong>and</strong> better flavour<br />
to the beer (Yeo, 2000). In the UK in 1998 the non-returnable market was split: cans<br />
69%, glass bottles 30%, plastic bottles (PET) 1%. Polyethylene terephthalate (PET) has<br />
been the preferred plastic used for containers of soft drinks since c. 1993. This product is<br />
gas permeable <strong>and</strong> carbon dioxide seeps out <strong>and</strong> oxygen seeps in. This has been<br />
overcome for beer by using a gas-proof coating on the PET or by the insertion of a nonpermeable<br />
barrier between two layers of PET (Nelson, 2000a). These bottles can be<br />
recycled <strong>and</strong> new bottles can be made of 40% recycled material. These new PET<br />
containers have greatly reduced oxygen ingress to the bottle <strong>and</strong> a shelf-life of six months<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
has been claimed for beer (Nelson, 2000a). Carbon dioxide escape from the bottle<br />
remains a problem. The most-used coatings involve a silicon oxide layer on the outside of<br />
the bottle, which means the beer is in contact only with the PET.<br />
There is now an alternative to PET, polyethylene naphthalate (PEN). This polymer has<br />
10±15 times greater barrier properties than PET (Nelson, 2000a). This gives the potential<br />
to refill the bottles as well as recycle the plastic. Bottles made of PEN can be washed at<br />
85 ëC (185 ëF) <strong>and</strong> so can be tunnel pasteurized containing the beer. PEN is much less gas<br />
permeable than PET, but is at least four times as expensive. A number of major brewers<br />
are now using PET <strong>and</strong> PEN bottles for beer, claiming the beer is easier to chill in these<br />
bottles <strong>and</strong> that drinking without a glass is also easier. There is no breakage in the supply<br />
chain <strong>and</strong> beer in plastic bottles is more acceptable at sporting events, where broken glass<br />
can be dangerous. There are likely to be further innovations in this area.<br />
21.4 Canning<br />
The arrangement of a canning line has much in common with that for non-returnable<br />
bottles (see above). This section will concentrate on the differences from bottling, <strong>and</strong><br />
will focus on the package itself <strong>and</strong> how it is filled. Beer, in cans, was first test-marketed<br />
in the USA in 1935. In 1962 the `rip-off' aluminium can end opening was introduced,<br />
again in the USA. Further developments occurred with various side seam techniques<br />
using tin-free steel on cans, which were at that time three-piece. The two-piece<br />
aluminium can was introduced in 1958, closely followed by other developments. The<br />
American domestic market has been dominated by beer in the can. In 1993 the market<br />
was split: can, 60%; non-returnable bottle, 25%; returnable bottle, 5%; draught, 10%.<br />
Other countries, particularly in Europe, concerned with the problem of empty can<br />
disposal, have not encouraged canning <strong>and</strong> the bottle dominates the market. However, the<br />
UK is unique among European countries <strong>and</strong> again the can is prevalent in the take-home<br />
trade. In 1998 the split of the market was: can, 24%; non-returnable bottle, 11%;<br />
returnable bottle, 2%; draught, 63%. The canning of beer, therefore, is clearly important<br />
in world terms but trends are certainly now towards bottles.<br />
The unique aspects of beer canning relate to the can itself, the unloading <strong>and</strong> rinsing of<br />
the cans prior to filing, the application of the date information <strong>and</strong> the filling <strong>and</strong> closing<br />
operations. All other aspects of palletizing <strong>and</strong> de-palletizing, secondary packaging <strong>and</strong><br />
pasteurization are very similar to the operations associated with bottling.<br />
21.4.1 The beer can<br />
Cans have some advantages to the brewer compared to bottles. They are lighter <strong>and</strong><br />
unbreakable. They can easily be stacked both in the brewery <strong>and</strong> in the refrigerator. The<br />
can has a large surface area on which to display information about the br<strong>and</strong> <strong>and</strong> it can<br />
be opened without a tool. Light cannot penetrate the can <strong>and</strong> so light-struck flavour<br />
cannot develop. Disadvantages relate almost entirely to disposal. Domestic disposal is<br />
usually to a waste tip mixed with other waste materials. This is not satisfactory.<br />
Attempts should be made to compress <strong>and</strong> recycle cans <strong>and</strong> to separate the two metals<br />
used in can manufacture, aluminium <strong>and</strong> steel. Making aluminium cans from aluminium<br />
ore takes considerable electrical energy. Only 10% of this energy is needed when new<br />
cans are made from recycled material. The need for active recycling programmes is<br />
clear.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Beer cans are now almost entirely of two-piece construction whether made from<br />
aluminium or steel. The can end, reflecting the original American invention, is always<br />
made of aluminium. The volume of the can is usually, 33, 44, or 50 cl. The metal used to<br />
make the can is largely governed by the conditions prevailing in the industries of the<br />
country of origin. Aluminium is almost always used in the USA whilst steel dominates in<br />
Germany. Both metals are used in the UK. Beer cans, both aluminium <strong>and</strong> steel, are<br />
characterized as 206 or 202 cans relating to the internal diameter of the mouth of the can.<br />
The newer 202 can was introduced to save metal <strong>and</strong> hence weight:<br />
diameters: 206, 57.4 Ô 0.3 mm; 202, 52.4 Ô 0.3 mm<br />
weights for 50 cl can (aluminium): 206, 19.32 g; 202, 18.38 g (Kunze, 1999).<br />
The tendency is for most beer canners to use the 202 can.<br />
The two-piece can is manufactured from a coil of plate, which is unwound, lubricated<br />
<strong>and</strong> fed into a blanking press (Scruggs, 1997), which forms the sheet into shallow cups.<br />
These cups proceed to an ironing press to form the side wall of the can. More metal is<br />
retained at the top <strong>and</strong> bottom of the can for strength but in the side wall the thickness<br />
will be only 0.09 mm. The subsequent strength of the filled can relies on the internal<br />
pressure of the beer in the can. Trimming follows ironing to ensure uniform height.<br />
During these processes the can is covered with lubricant that is removed with sprays.<br />
External decoration on the can is applied by a rotary machine using inks <strong>and</strong>/or varnishes,<br />
which are stabilized by baking. Internal coatings are applied by an airless stationary<br />
spray-gun <strong>and</strong> the cans are then baked again to cure the coating <strong>and</strong> remove solvent. The<br />
final process is the die necking to achieve the 206 or 202 diameter. The cans are supplied<br />
to the brewery on pallets. For 33 cl cans 8,280 cans can be put on one pallet in 23 layers.<br />
For canning lines operating at 2,000 cans/min. the constant supply of cans to the canning<br />
line is critical for success. Manufacturers of cans often have plants close to major<br />
brewery sites.<br />
Can ends are made from 0.27 mm thick aluminium sheet. The end contains the ringpull,<br />
which nowadays is almost entirely of the `stay-on-tab' type which means that<br />
separated, <strong>and</strong> discarded rings do not pollute the environment. Lids are supplied with a<br />
diameter of 64.75 mm which, after fitting to the body of the can, is reduced to the<br />
appropriate size for the 206 or 202 can.<br />
21.4.2 Preparing cans at the brewery for filling<br />
The main concern with h<strong>and</strong>ling cans, compared to bottles, is the delicate nature of the<br />
can <strong>and</strong> that it arrives pre-decorated. This decoration must not be damaged, as it cannot<br />
be repaired. Cans are pushed off the pallets in layers onto a flat conveyor chain. There<br />
must be no gaps between the cans so they cannot fall over. The spacer packaging <strong>and</strong> the<br />
steel frame of the pallet is normally returned to the supplier. Can lids are also supplied on<br />
pallets, which can contain 250,000 lids packed in bags of 600. The `best-by' date is<br />
usually displayed on the base of the can. This is usually applied while the can is dry, i.e.,<br />
before rinsing <strong>and</strong> filling. The normal method is to use an ink-jet printer, which can<br />
match the speed of can filling. Cans are always rinsed before filling by spraying with<br />
water whilst in the inverted position. The motive force for this act is usually gravity <strong>and</strong><br />
hence cans are unloaded at a height of 3 m <strong>and</strong> then pass down through the rinser to the<br />
filler. Cans are delicate <strong>and</strong> have little structural strength when empty. They will not,<br />
however, break like glass bottles <strong>and</strong> so concerns about glass fragments dem<strong>and</strong>ing<br />
complex inspection machinery are not relevant to canning.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
21.4.3 Can filling<br />
The most important machine on the canning line is the filler. There are similarities in the<br />
machines used for filling bottles <strong>and</strong> cans. Some machines are dual purpose <strong>and</strong> have the<br />
change parts to fill bottles <strong>and</strong> cans. Dedicated machines are used for choice because the<br />
speeds at which a can filler can run are about twice those of a bottle filler. A 100 head can<br />
filler will operate at least at 2,000 cans per minute. The main technical differences<br />
between bottle <strong>and</strong> can filling derive from the properties <strong>and</strong> size of the empty can. Thus,<br />
cans have a very wide neck <strong>and</strong> so provide a large surface area for gaseous exchange<br />
during filling. This can lead to a loss of carbon dioxide <strong>and</strong> ingress of oxygen <strong>and</strong> cans<br />
are very light <strong>and</strong> will not resist a vacuum so they are not evacuated during filling. The<br />
filling tube has similarities with the short bottle filling tube. A quiet fill down the side of<br />
the can is achieved by lowering the filling assembly to achieve an air-tight connection<br />
with the can, i.e., the can is not lifted in this process.<br />
The technique of centring the delicate can is important. Some very clever designs<br />
allow operation at high speed (> 1,500 cans/min.) with no oxygen entry. Between the<br />
lowering centring tulip <strong>and</strong> the fixed filling tube is a differential pressure chamber that<br />
connects to the internal space of the can. During counter-pressuring <strong>and</strong> filling this<br />
chamber maintains the filling pressure. There is only a small difference between the<br />
diameter of the pressure chamber <strong>and</strong> the diameter of the can <strong>and</strong> so there is only a small<br />
force on the can during filling. After filling the volume in the differential pressure<br />
chamber is reduced <strong>and</strong> pressure increases <strong>and</strong> the can is thus released from the seal.<br />
The beer is held in an annular filling bowl under a blanket of carbon dioxide. The<br />
filling valve is supported by the annular bowl (Fig. 21.11). The sequence of filling is:<br />
· carbon dioxide purging<br />
· counter-pressuring<br />
· filling<br />
· venting<br />
· lifting the valve.<br />
Stationary<br />
neutral cam<br />
Rotation<br />
Stationary<br />
closing cam Stationary<br />
snift cam<br />
Rotating<br />
filler bowl<br />
Infeed<br />
Valve tulip<br />
operating cam Rotation<br />
Diverter<br />
guide<br />
Purge cam<br />
Infeed conveyor<br />
Air operated<br />
vacuum cam<br />
Air operated<br />
valve trip<br />
Infeed worm<br />
Infeed star<br />
Discharge<br />
conveyor<br />
To<br />
seamer<br />
Fig. 21.11 Can filling, schematic. Cans are fed to the infeed star, which places them on to the<br />
filler bowl platforms, one can beneath each filling valve. The can filling sequence then operates;<br />
as the bowl rotates, cams remove air from the can, charge with gas, fill with beer, close the<br />
valve <strong>and</strong> remove residual gas. (Heins <strong>and</strong> Heuer, 1997).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
During filling the filling valve is lowered along with the filling tube, the return air-pipe<br />
<strong>and</strong> the differential pressure chamber. The airtight connection is so formed. The can is<br />
then counter-pressured with carbon dioxide <strong>and</strong> the beer valve opens introducing the beer<br />
quietly down the side of the can from, usually, 15 tubes. In some can fillers these small<br />
tubesarereplacedbyavalve,whichdeliversthebeerinathinfilmdownthewholeofthe<br />
can side to the bottom. The evacuated air flows up the return air pipe until the beer level<br />
reaches the pipe. The pipe contains aball, which rises with the beer <strong>and</strong> closes the inlet.<br />
This controls the height of liquid in the can as with bottle filling. The position of the<br />
return air-pipe can be adjusted. The centring tulip is then raised, the differential pressure<br />
is increased, <strong>and</strong> the can is released from the seal <strong>and</strong> proceeds to the can seamer for<br />
closing. There is little fob in this process <strong>and</strong> it is not economic to attempt recovery from<br />
the evacuated air. In some systems after filling, jetting carbon dioxide over the beer<br />
surface breaks any bubbles of carbon dioxide or air.<br />
Controllingfillingheightbythepositionofthereturnair-pipeisanestablishedprocedure<br />
for bottles <strong>and</strong> cans. This does, though, require slowing the filling process towards the end.<br />
For this reason there has been adesire to develop procedures of volumetric fill where the<br />
exact filling volume is pre-determined (Fig. 21.12). Avolumetric can filling system has<br />
been described (Kunze, 1999) in which the volume is measured in a thin measuring cylinder<br />
by a floating ultrasonic sensor. The cylinder is filled from below without turbulence <strong>and</strong> is<br />
hence charged to deliver the appropriate volume at the beginning of the filling process. This<br />
allows a speeding <strong>and</strong> st<strong>and</strong>ardizing of the filling process.<br />
21.4.4 Can closing (seaming)<br />
There is a large surface of beer in the can <strong>and</strong> so the lid must be secured as soon as<br />
possible to keep out oxygen <strong>and</strong> to minimize dissolved oxygen levels ( 0.2 mg/l). Cans<br />
must be hermetically sealed to be impervious to air, liquids <strong>and</strong> bacteria. The closed can<br />
must withst<strong>and</strong> the internal pressure <strong>and</strong> be robust enough to survive distribution <strong>and</strong><br />
retail display.<br />
The can lid is usually placed on the can whilst the can is on the filler carousel. The can<br />
is raised <strong>and</strong> pressed together with the loosely placed lid against a firm sealing head. A<br />
double seaming process follows.<br />
· First operation, interlocking: the outer part of the lid is bent under by a pre-roller with<br />
a defined profile.<br />
· Second operation, tightening: an air-tight seal is obtained by pressing with a seaming<br />
roller.<br />
Carbon dioxide is blown over the surface of the beer as the can end is being slid into<br />
position. Most of the air is blown out <strong>and</strong> replaced with carbon dioxide (under-cover<br />
gassing). Air contents in the head space are < 1.5 ml. A sealing compound incorporated<br />
in the curl end of the lid effects the hermetic seal following tightening. The compound<br />
may be water-based latex emulsion or synthetic rubber material.<br />
The success of the seaming operation depends on smooth coordinated operation of all<br />
moving parts of the seaming machine. Regular, planned maintenance is essential guided<br />
by the history of the machine. There should be regular corrective action to ensure the<br />
machine is always set precisely for perfect closing of the can. An efficiency of operation<br />
of 95% (actual time run versus time planned) should be achievable through proactive<br />
planned maintenance. This type of maintenance is best achieved through the team<br />
approach with multi-disciplined teams of operators able to work together on the line,<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Cam follower for<br />
valve lowering<br />
Differential<br />
pressure<br />
chamber<br />
Filling valve<br />
Centring unit<br />
Level probe<br />
building up a history of operation <strong>and</strong> utilizing this knowledge to effect the maintenance<br />
plan. Programmes of regular testing for seam integrity should be in place. St<strong>and</strong>ard visual<br />
tests have been developed for this as well as destructive tests involving examination of<br />
seam tolerances with micrometer gauges (Heuer, 1997).<br />
Since cans are not transparent fill height is usually checked using radiation which<br />
penetrates the can <strong>and</strong> checks whether the beer fill height meets specification. Finally<br />
almost all canned beer is tunnel pasteurized. Typically beer will receive 20 PU:<br />
· pre-heat 46 ëC (115 ëF) 1 min.<br />
· superheat 62 ëC (144 ëF) 3±4 min.<br />
· pasteurize 60 ëC (140 ëF) 15 min.<br />
· pre-cool 46 ëC (115 ëF) 2 min.<br />
· cool 32 ëC (90 ëF) 2 min.<br />
Product channel<br />
Automatically<br />
applicable CIP cup<br />
1 Flushing, pressurizing<br />
<strong>and</strong> return gas valve<br />
2 Upper snift valve<br />
3 Lower snift valve<br />
4 Flushing <strong>and</strong> CIP<br />
return valve<br />
Metering chamber<br />
Beer supply valve<br />
Following pasteurization the cans are dried with jets of compressed air <strong>and</strong> the fill height<br />
is often checked again by radiation. Secondary packaging then takes place often to the<br />
customer's specification.<br />
1<br />
2<br />
3<br />
4<br />
Snift channel<br />
Pressurizing channel<br />
Fig. 21.12 Can filling valve (by courtesy of Krones).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Microbiological checks focus on the beer supply prior to filling <strong>and</strong> the filling <strong>and</strong><br />
seaming operations.<br />
21.4.5 Widgets in cans<br />
Traditional English ales were, <strong>and</strong> sometimes still are fermented in open vessels. The<br />
naturallevelofcarbondioxidethusentrainedinbeerattheendoffermentationwasabout<br />
1.2 vol/vol or 2.4g/l (Lindsay et al., 1995). This was therefore the normal carbon dioxide<br />
level at which the ale was drunk. Canned ale was frequently packaged with acarbon<br />
dioxide content of 2.6 to 3.0 vol/vol. This obviously had a very different taste to<br />
traditional cask ale. The beer was much more effervescent <strong>and</strong> more like alager beer in<br />
mouth-feel. It had been discovered in the 1940s that small amounts of nitrogen gas in<br />
beer had aprofound effect on head creaminess <strong>and</strong> durability (Carrol, 1979). This was<br />
partly as a result of the low solubility of nitrogen gas in beer (20mg/l at room<br />
temperature <strong>and</strong> pressure). Nitrogen is colourless, tasteless, odourless <strong>and</strong> chemically<br />
inert. These properties <strong>and</strong> discoveries suggested the use of nitrogen in the packaging of<br />
ale with the objective of providing the drinker with ataste sensation akin to that of cask<br />
beer, i.e., low carbon dioxide effervescence <strong>and</strong> thick creamy head. The problem was to<br />
introduce nitrogen into the beer in areproducible way that could then consistently give<br />
the consumer the appropriate taste sensation. Considerable research <strong>and</strong> development<br />
work carried out in the early 1980s led to the development of the `widget'.<br />
The concept was to introduce aplastic capsule with asmall hole <strong>and</strong> containing<br />
nitrogen into the can <strong>and</strong> to pressurize it during canning. The pressure would be released<br />
when the can was opened <strong>and</strong> the nitrogen would be forced into the beer giving the<br />
characteristic creamy head <strong>and</strong> reflux of bubble formation (Lindsay et al., 1995). Many<br />
widgets have been described (Brown, 1997). Some have been more successful than<br />
others. Originally widgets were inserted into the can before filling but this did not allow<br />
the exclusion of oxygen <strong>and</strong> hence flavour stability was poor. Now widgets can be<br />
supplied already attached to the base of the can. Frequently the floating widget is used<br />
since it is less likely to trap oxygen in the can during filling. Liquid nitrogen is often<br />
added during filling. Before the development of the widget some companies used solely<br />
liquid nitrogen to promote the reflux effect. This technique is still used in the keg<br />
packaging of `smooth flow' ales. In any event oxygen ingress must be kept as low as<br />
possible during canning <strong>and</strong> nitrogen can be used as the undercover gas. Cans often<br />
proceed through atunnel between the filler <strong>and</strong> the seamer to further prevent oxygen<br />
pick-up (Brown, 1997).<br />
Many companies introduced smooth flow ales in cans in the 1980s <strong>and</strong> 1990s. Some<br />
thought that this type of beer would replace st<strong>and</strong>ard canned ales with carbon dioxide<br />
contents of around 2.5 vol/vol. This has not happened <strong>and</strong> many consumers returned to<br />
the more effervescent product.<br />
21.5 Kegging<br />
Sales of beer in draught form are greater than in small-pack in only the UK <strong>and</strong> Irel<strong>and</strong><br />
(Table 21.1). In these countries the sale of beer in kegs is important. Kegging is about<br />
filling carbonated, pressurized, pasteurized beer into sterile containers. These containers<br />
usually contain 25, 30, 50, or 100 litres of beer. In the UK volumes of 9 imp. gal. (firkin),<br />
18 imp. gal. (kilderkin or kil), or 36 imp. gal. (barrel) are still common. All kegs are<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
eturnable. The collection of empty kegs from depots, bars <strong>and</strong> public houses is an<br />
important part of the overall management of keg packaging (Chapter 22). Kegging has<br />
similarities with the packaging of beer in returnable bottles. The major differences<br />
concern the h<strong>and</strong>ling <strong>and</strong> cleaning of the much bigger container <strong>and</strong> the sterilization of<br />
the beer. As noted beer inkegs cannot be tunnelpasteurized. The theoryof pasteurization<br />
(Section 21.3.2) shows how beer can be pasteurized in bulk.<br />
21.5.1 The keg<br />
Kegs are normally made of stainless steel or aluminium. Stainless steel kegs are made of<br />
achrome/nickelalloy. Theyareheavy;a50litrekegwillweigh 12±15kg(around30lb.).<br />
Aluminiumkegs aremadeofanalloyalsocontaining magnesium <strong>and</strong>silicon.Thesekegs<br />
are lighter than stainless steel kegs <strong>and</strong> were originally more popular. However<br />
aluminium kegs are more frequently stolen than stainless steel kegs because of the ease<br />
with which aluminium can be melted down <strong>and</strong> sold. Aluminium kegs cannot be cleaned<br />
with caustic alkali-based detergents because hydrogen gas is formed. Cleaning is with<br />
acid or dilute alkalis. Stainless steel kegs can be cleaned with acid or alkaline detergents<br />
<strong>and</strong> generally are more robust in use, aproperty that is particularly important as the<br />
container ages. Generally, therefore, stainless steel kegs are preferred.<br />
All kegs have aneck containing athreaded bush (Barnes neck) into which fits akeg<br />
valve fitting (Fig. 21.13). This fitting is called the spear or extractor tube <strong>and</strong> through it<br />
the filling, emptying, cleaning <strong>and</strong> automatic closing of the keg is achieved. Kegs have<br />
advantages over bottles. They allow the partial dispense of the product <strong>and</strong> they operate<br />
as closed vessels with in-built leakage detection. Keg extractors should not be withdrawn<br />
outside the brewery <strong>and</strong> kegs are returned containing excess gas pressure, which prevents<br />
contamination entering. On modern keg filling lines there is a pressure test to<br />
demonstrate internal pressure <strong>and</strong> any kegs not having such apressure will be rejected<br />
<strong>and</strong> not filled.<br />
Kegs are delivered with the extractor installed <strong>and</strong> protected from dirt during delivery<br />
withaplastickegcap.Todispensethebeerabayonet-typedispenseheadisclampedontothe<br />
extractoratthebar.Thisallowstheingressofthetoppressuregas<strong>and</strong>theoutletofthebeerto<br />
the dispense tap (Chapter 23). There are several types of fitting which means that kegs from<br />
different brewers are often not interchangeable on the dispense equipment in the bar.<br />
Protective<br />
top chime<br />
472<br />
∅ 382<br />
Protective<br />
bottom chime<br />
Spear for<br />
filling <strong>and</strong><br />
emptying<br />
under top<br />
pressure<br />
of CO 2<br />
Fig. 21.13 Vertical section of a50l beer keg, height 472mm; diameter 382mm (by courtesy of<br />
Alumasc Ltd).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Well-fittings contain two valves, one for the beer, normally aball valve <strong>and</strong> one for<br />
the top pressure gas. Aflat-top fitting has one valve which seals separately the gas <strong>and</strong><br />
the beer. It is easy to use but is bigger than the well-fitting. The `combi' fitting is a<br />
combination of the well <strong>and</strong> flat top fitting, but has two valves. Great efforts have been<br />
made to make these fittings safe <strong>and</strong> tamper-proof. It is now an integral part of the filling<br />
operation to test for extractor tightness as well as internal pressure in the keg before<br />
filling.<br />
21.5.2 Treatment of beer for kegging<br />
Beer for kegging is normally conditioned to yield 1.5 to 2.5 vol/vol carbon dioxide.<br />
This applies to ale <strong>and</strong> lager (Chapter 15). Some `smooth-flow' ales in the UK are<br />
now packaged at carbon dioxide levels of 1.0 volume <strong>and</strong> with nitrogen contents of<br />
30±40mg/l. Nearly all beer for kegging is bulk (flash) pasteurized in acontinuous<br />
flow pasteurizer at high pressure, say 10 bar (150lb./in. 2 ),against aback pressure of<br />
1bar (15lb./in. 2 ).<br />
Flash pasteurization<br />
Flash pasteurization is carried out in aplate heat exchanger in which there are four<br />
sections (Fig. 21.14):<br />
· regeneration section<br />
· heating section<br />
· holding tube<br />
· cooling section.<br />
Beer is pumped to the regeneration section where it flows counter-current to hot beer <strong>and</strong><br />
is therefore pre-heated. In the heating section it is brought up to pasteurization<br />
temperature by passing counter-current to hot water or steam. It is then held for apredetermined<br />
period in the holding tube. The beer then passes back to the regeneration<br />
section where it loses heat to the incoming beer. It subsequently runs counter-current to<br />
cold brine or alcohol in the cooling section. The maximum temperature achieved is<br />
between 71 <strong>and</strong> 79ëC (160±175ëF) <strong>and</strong> the holding period is usually between 15 <strong>and</strong> 60<br />
seconds.<br />
It is very important that all the beer flowing through the heat exchanger receives the<br />
same pasteurization treatment. Turbulent flow ensures that this is the case. The Reynolds<br />
number can be used to ensure that the condition of turbulent flow is achieved. This<br />
number is the product of liquid density, velocity <strong>and</strong> tube diameter divided by liquid<br />
viscosity. At Reynolds number values below 2,000 flow is laminar but above 3,000 flow<br />
is increasingly turbulent <strong>and</strong> this should be aimed for.<br />
To change the number of pasteurization units given to beer the temperature is altered.<br />
The heat exchanger is designed for a particular flow rate for maximum efficiency. There is<br />
a substantial pressure drop through the pasteurizer <strong>and</strong> to keep carbon dioxide in solution<br />
beer is pumped in at 8.5 to 10 bar gauge pressure against a back pressure of 1 bar. The use<br />
of buffer tanks before <strong>and</strong> after the pasteurizer is essential to prevent interruptions of flow<br />
<strong>and</strong> pressure surges on the beer in the bright beer tanks <strong>and</strong> the keg racker.<br />
Heat transfer in the exchanger is achieved mainly by convection. An important design<br />
factor is the surface area required for efficient heat transfer:<br />
A ˆ Q/HT<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Zephyr valve<br />
To filler<br />
or racker<br />
Rhodes flow indicator<br />
Zephyr valve<br />
Drain<br />
Heater<br />
Detergent–sterilizing tank<br />
Steam<br />
supply<br />
Drain<br />
Paraflow<br />
b<br />
Brine/<br />
alcohol<br />
Cooling<br />
Beer pump<br />
Control<br />
cock<br />
Duplex<br />
temperature<br />
recorder<br />
Regeneration Heating Holding<br />
Pump<br />
Temperature<br />
controller<br />
Hot water set<br />
b a<br />
Oil-free air supply<br />
Diaphragm<br />
valve<br />
Steam<br />
reducing<br />
valve<br />
Steam<br />
supply<br />
Steam<br />
injector<br />
Fig. 21.14 Flash pasteuriser; (a) raw unpasteurized beer, (b) pasteurized beer (by courtesy of APV Co. Ltd).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Unpasteurized<br />
carbonated<br />
beer tank
where Qis the heat load in joules/h, Ais the area in m 2 ,His the overall heat transfer<br />
coefficient in watts/m 2 /K, <strong>and</strong> Tis the logarithmic temperature difference (Hough et al.,<br />
1982).<br />
Advantages of flash pasteurization compared to tunnel pasteurization are:<br />
· less space required<br />
· lower capital cost of equipment<br />
· lower operating costs (only 15% of the cost of tunnel pasteurization (Hyde, 2000))<br />
· shorter periods of exposure of the beer to temperatures where chemical changes are<br />
rapid but pasteurization is slow.<br />
There are dangers for flavour stability of beer with flash pasteurization. If dissolved<br />
oxygen levels are >0.3mg/l more marked flavour changes may occur in flash<br />
pasteurization than in tunnel pasteurization. This is owing to higher temperatures <strong>and</strong><br />
turbulent flow but more importantly, from the recycling of beer back to the buffer tank<br />
when the process conditions have not been met or packaging has been interrupted. This<br />
leadstoexcessivepasteurization.Itisgood<strong>practice</strong>tocirculatewater,<strong>and</strong>notbeer,when<br />
flow is stopped but this is difficult to control <strong>and</strong> can lead to beer losses <strong>and</strong> dilution.<br />
Asthebeerissterilizedbeforefilling,thekegsthatreceivethebeermustbesterile<strong>and</strong><br />
filling must be an aseptic operation. Strict microbiological checks downstream of the<br />
pasteurizer are essential. Atypical pasteurizer may hold the beer for 20 seconds at 75ëC<br />
(167ëF) <strong>and</strong> this is equivalent to 50PU. This high PU value illustrates the need for close<br />
control of the pasteurizer to avoid excess pasteurization <strong>and</strong> consequent flavour<br />
deterioration of the beer.<br />
21.5.3 H<strong>and</strong>ling of kegs<br />
Kegs are transported in stacks, which can be arranged horizontally using pallets or<br />
vertically using spacer boards. Horizontally configured cradle pallets are very heavy <strong>and</strong><br />
contribute considerably to the overall weight on adistribution vehicle. Plastic spacer<br />
boards are light <strong>and</strong> allow the vertical stacking of containers up to five stacks high. This<br />
system is now preferred in the UK <strong>and</strong> Irel<strong>and</strong>.<br />
Empty kegs returning to the brewery first have to be destacked or depalletized. The<br />
principleofoperation ofthesemachinesresemblesthatofthe machines usedfor h<strong>and</strong>ling<br />
crates of returned bottles (Section 21.3.1). Destacking is performed layer by layer by<br />
pushing the kegs together <strong>and</strong> lifting by pneumatic grippers. The spacer board can be<br />
removed manually or by machine. Stacking the full containers at the end of the line <strong>and</strong><br />
after labelling is performed by similar machines. The spacer boards can be placed in the<br />
stacks manually or by machine.<br />
Efficiency of operation is critical to any kegging line. It depends as much on the<br />
supply of the containers as it does on the supply of the beer. It is essential that kegs are<br />
`fit to fill' <strong>and</strong> that kegs that are not fit do not proceed to the racking machine<br />
(International Bottler <strong>and</strong> Packer, 2000). After destacking, empty kegs are tested for<br />
internal pressure <strong>and</strong> tightness of the extractor tube in the Barnes neck. Kegs that fail<br />
are removed from the line for subsequent inspection <strong>and</strong> repair. These kegs may have<br />
been tampered with <strong>and</strong> thus be heavily infected or unsafe. Plastic protective caps will<br />
have been removed at the dispense site. The next requirement is to wash the exterior of<br />
the keg. Most returned kegs display a self-adhesive label, which will show the details<br />
of the previous filling, i.e., the beer quality <strong>and</strong> `best before' data. These old labels<br />
must be removed. This is achieved by pre-soaking, scrubbing <strong>and</strong> spraying with hot<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
detergent at 70ëC (160ëF). Detergent is re-circulated. Hot water sprays remove<br />
detergent before the kegs leave the washer. External keg washers take up alot of space<br />
<strong>and</strong> use much energy <strong>and</strong> water (at least 10 hl/hourat 1.4 bar). The supply ofhot water<br />
is frequently waste process water <strong>and</strong> condensate, which may already have been used<br />
several times in the brewery. Roller or flat top chain conveyors convey kegs through<br />
these machines.<br />
21.5.4 Keg internal cleaning <strong>and</strong> filling<br />
Kegs are filled with pasteurized beer. Kegs must, therefore, be sterile as well as clean<br />
prior to receiving the beer. For this reason cleaning, sterilizing <strong>and</strong> filling are carried out<br />
on one machine called akeg racker. The extractor (spear) remains in place during this<br />
operation. On most machines the cleaning takes place with the keg inverted, when the<br />
drainage is quicker <strong>and</strong> more complete <strong>and</strong> water forced up the spear cleans the side <strong>and</strong><br />
base of the keg more effectively. In most systems filling also takes place with kegs<br />
inverted but any misalignment of the filler <strong>and</strong> the neck can lead to large beer losses <strong>and</strong><br />
so sometimes each keg is returned to the upright position for filling (Fig. 21.15). These<br />
tasks dem<strong>and</strong> the inclusion in the line of 180ë turning machines, which must operate<br />
reliably <strong>and</strong> be of robust construction.<br />
Automatic keg cleaning <strong>and</strong> filling machines are generally of two basic types, inline<br />
or rotary. In-line machines have anumber of `lanes' at which cleaning, sterilizing<br />
<strong>and</strong> filling takes place in sequence at aseries of `heads'. The number of lanes (typically<br />
8±24) governs the capacity of the machine. The lanes are at right-angles to the<br />
conveyor bringing the empty kegs to be filled. In-line machines are often used in the<br />
UK <strong>and</strong> they have to cope with three sizes of keg, e.g., 50l(11 imp. gal.), 18 imp. gal.<br />
<strong>and</strong> 36 imp. gal. In these machines kegs are fed from one side <strong>and</strong> leave from the other<br />
<strong>and</strong> proceed in astraight line to the labeller <strong>and</strong> capper (Figs 21.15a <strong>and</strong> b). These<br />
diagrams show a typical overall layout <strong>and</strong> the detail of the cleaning <strong>and</strong> filling<br />
operation. Asimple two-head machine is shown at which washing takes place on the<br />
first head <strong>and</strong> filling on the second. Recently machines with four heads have been<br />
developed to provide alonger cleaning cycle (Carter, 2001). These machines have two<br />
washer heads, a sterilization station, a pre-filling head <strong>and</strong> a filler head. Stringent<br />
washing using water <strong>and</strong> detergent can be carried out. In-line machines can h<strong>and</strong>le<br />
frequent size changes <strong>and</strong> short runs of different beer qualities, though this will<br />
considerably reduce efficiency. Downtime on an individual lane will not affect the<br />
other lanes <strong>and</strong> individual lanes can be taken out for maintenance during normal<br />
production. Lanes can also be added to an existing machine to increase production.<br />
Production rates of 1,200 kegs/h are achievable.<br />
On rotary machines kegs are h<strong>and</strong>led simultaneously on aseries of stations in a<br />
circular motion around a central core. A typical rotary machine would comprise a<br />
washing machine with 24 stations <strong>and</strong> afilling machine with 12 or perhaps 16 stations<br />
(Fig.21.16a).Rotationtimesare65±125seconds.Kegsarebroachedonce<strong>and</strong>thenrotate<br />
with services switching on <strong>and</strong> off as the kegs move around the circle. Services of water,<br />
detergent,steam,carbondioxidegas<strong>and</strong>beer<strong>and</strong>condensatereturnhavetobeconnected<br />
into the centre of the machine. This requires complex sealing with two discs to prevent<br />
mixing (Figs 21.16b <strong>and</strong> c). These machines require less space than in-line machines <strong>and</strong><br />
are capable of production rates of up to 2,000 kegs/h.<br />
The mechanical design of rotary machines is simpler than in-line machines <strong>and</strong><br />
downtime is usually less, however, if one part of the machine fails the whole line is<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
(b)<br />
(a)<br />
Lateral<br />
discharge<br />
conveyor<br />
Depalletizer<br />
Locator<br />
boards<br />
Unitizer<br />
Decant<br />
Capper<br />
Filler<br />
head<br />
Racking<br />
(filling)<br />
station<br />
Fit for<br />
fill<br />
Labeller<br />
Control<br />
systems<br />
Manifolds for<br />
steam, water,<br />
detergent,<br />
beer, etc.<br />
Empties<br />
turner<br />
Tipping<br />
station<br />
Washing<br />
station<br />
Washer<br />
head<br />
Inverted<br />
keg<br />
Lateral<br />
entry<br />
conveyor<br />
Keg<br />
positioning station<br />
External washer<br />
Washer/racker<br />
Checkweigher<br />
Container flow<br />
Fulls<br />
turner<br />
Conveyor for<br />
re-circulating<br />
Fob<br />
tank<br />
Sterile<br />
beer<br />
tank<br />
Pasteurizer<br />
Buffer<br />
tank<br />
Beer<br />
flow<br />
Beer<br />
ex BBT<br />
Fig. 21.15 Automatic linear (lane) internal keg washing <strong>and</strong> filling machine, equipment shown<br />
in (a) is found in the washer/racker area of (b) (Eaton, 2002; Hough et. al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
stopped. Rotary machines are not suitable for frequent container size <strong>and</strong> beer quality<br />
changes. In the UK the cleaning agent is usually acid (e.g. 2% phosphoric acid). This<br />
allows mixed populations of aluminium <strong>and</strong> stainless steel kegs to be processed on the<br />
same machine. Constant use of acid will lead to build up of protein residues on the inside<br />
of the keg <strong>and</strong> from time to time kegs should be sorted <strong>and</strong> stainless steel kegs cleaned<br />
with caustic alkali. Aluminium kegs can be cleaned with dilute alkali. On the mainl<strong>and</strong> of<br />
Europe caustic soda is normally used as stainless steel kegs predominate. On some<br />
machines there is the facility to use both acid <strong>and</strong> alkali in the same cleaning sequence.<br />
This is not common <strong>practice</strong> in the UK.<br />
Table 21.2 Time sequences for automatic cleaning, sterilizing <strong>and</strong> filling of 50 l <strong>and</strong> 100l kegs on<br />
a four-head racking machine (Carter, 2001).<br />
Sequence operation Time (s) 50 litre Time (s) 100 litre<br />
Head 1<br />
Deullage 1<br />
3 5<br />
Wash 1, recovered water 70 ëC 8 15<br />
Air purge 6 12<br />
Wash 2, detergent 18 28<br />
Air purge 7 15<br />
Vent head 1 1<br />
Transfer delays 12 14<br />
Time at head 1 55 90<br />
Head 2<br />
Wash 3, water 70 ëC 20 28<br />
Steam purge 105 ëC 7 17<br />
Steam scavenge 8 16<br />
Steam pressurize 1.5 bar 2 3<br />
Head cool 1 1<br />
Transfer delays 17 25<br />
Time at head 2 55 90<br />
Steam hold station<br />
Steam hold 46 81<br />
Transfer to head 3 9 9<br />
Time at steam hold station 55 90<br />
Head 3 (pre-filler)<br />
Steam hold 28 28<br />
Steam head 3 3<br />
CO2 purge 8 15<br />
CO2 pressurize 2 bar 2 3<br />
Vent head 2 2<br />
Transfer delays 12 39<br />
Time at head 3 55 90<br />
Head 4<br />
Steam head 3 3<br />
Beer fill 4±5 bar 35 70<br />
Scavenge 3 3<br />
Neck wash 3 3<br />
Transfer delays 11 11<br />
Time at head 4 55 90<br />
1 The process of allowing spent beer in the keg to drain.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Each C-frame<br />
incorporates the 11<br />
components for<br />
two treatment 10<br />
stations<br />
9<br />
8<br />
12<br />
13<br />
14<br />
Starwheels are<br />
exchangeable to<br />
suit kegs of<br />
different diameters<br />
Circular<br />
motion<br />
Sequences of cleaning, sterilizing <strong>and</strong> filling vary considerably in <strong>practice</strong> <strong>and</strong> for in-line<br />
machineswilldependonthenumberofheadsonthelane.Atypicalsequenceforafour-head<br />
in-line machine is shown in Table 21.2. The washes can be pulsed to give more reliable<br />
cleaning <strong>and</strong> each wash except the final one is purged from the keg with sterile air. The final<br />
wash is purged with steam. Effective sterilization <strong>and</strong> prevention of oxygen pick-up is<br />
essential. Steam is used for sterilization. Counter-pressuring with an inert gas such as carbon<br />
dioxide is used to keep out oxygen. A temperature of 105 ëC (190 ëF) must be recorded inside<br />
the keg for it to be effectively sterilized. Counter-pressures can be varied but are usually in the<br />
range 0.7±3.5 bar (10±50 lb./in. 2 ). Gas is removed down the spear as filling proceeds through<br />
the gas ports of the keg. The beer flow is modulated to avoid fobbing, starting <strong>and</strong> ending at<br />
about 10% of full flow. This procedure also allows more precise volume control (see below).<br />
The sequences (Table 21.2) demonstrate the longer time for processing the larger keg.<br />
The time sequence will be shorter on a two-head machine but the cleaning <strong>and</strong><br />
sterilization may not be as effective (Hough et al., 1982). It is advantageous to process<br />
long runs of beer into one keg size; frequent changes of size during a shift must be<br />
1<br />
2 3<br />
Fig. 21.16(a) A rotary keg racking machine shown in plan from the top. Kegs move around<br />
the central core at the series of stations which comprise: 1 cylinder down; 2 leak test; 3 rinse<br />
head; 4 blow out head; 5 pressure test; 6 counter-pressure; 7 fill; 8 disconnect; 9 rinse head; 10<br />
pressure relief; 11 cylinder up; 12 close keg; 13 release keg clamp; 14 discharge keg (Eaton,<br />
2002 <strong>and</strong> courtesy of KHS Till).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
7<br />
4<br />
5<br />
6
Fig. 21.16(b) <strong>and</strong> (c) Central distributor of the rotary keg racking machine shown as (b) a<br />
schematic block diagram <strong>and</strong> (c) in vertical section. Each seal for the supply <strong>and</strong> return of services<br />
comprises two discs: a carbon ring <strong>and</strong> a metal distributor ring. When the holes of the two<br />
rings in the rotating section are aligned liquid will flow to the head (Courtesy of KHS Till <strong>and</strong><br />
Eaton, 2002).<br />
avoided. Washer heads <strong>and</strong> filler heads have manifolds for detergent, water, steam,<br />
carbon dioxide <strong>and</strong> beer. All these mains must be cleaned <strong>and</strong> steam sterilized before beer<br />
is passed through. Keg plants require frequent cleaning. This can take up to six hours<br />
twice a week depending on the volumes of beer being processed. After filling, the Barnes<br />
neck is usually air-dried, though this step is sometimes omitted, <strong>and</strong> the keg proceeds to<br />
be capped <strong>and</strong> labelled. These tasks were traditionally manual but machines are now<br />
available for automatic capping <strong>and</strong> labelling.<br />
A major factor in filling is an accurate determination of the contents of the keg. This is<br />
required by statute in most countries <strong>and</strong> is subject to audit. There are two requirements:<br />
the contents must meet a prescribed amount so that the customer is assured of receiving<br />
the appropriate volume <strong>and</strong> is not defrauded, <strong>and</strong> containers must not be consistently<br />
overfilled or the correct amount of excise duty will not be paid. Control of volume to<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
meet these criteria is not easy. A gross <strong>and</strong> tare weighing system can be used with success<br />
on smaller volumes but this system has been unsuccessful with barrels (36 imp. gal.<br />
containers) still in use in the UK. A volumetric control is preferred although recently<br />
other devices have been described (Carter, 1998; Brewer <strong>and</strong> Carter, 2000).<br />
An on/off beer filling valve can give a keg brim fill to Ô0.5%. This can be improved to<br />
Ô0.25% with an electro-magnetic flow meter (Carter, 1998). More advanced systems<br />
(Brewer <strong>and</strong> Carter, 2000) incorporate a modulating back pressure valve with the flow<br />
meter. This system delivers pressure profile filling with a precision of Ô0.02%. No<br />
mechanical valve is required to regulate flow <strong>and</strong> 100 l/min. is achievable. A slow initial<br />
fill is possible, which reduces carbon dioxide breakout <strong>and</strong> the potential for oxygen pickup.<br />
After the speeding-up of the flow a quiet cut-off ensures low carry-over into the<br />
extractor tube <strong>and</strong> the retention of internal pressure. Other developments incorporate an<br />
inductive flow meter in a direct flow control filling head (International Bottler <strong>and</strong><br />
Packer, 2000). The keg is counter-pressured with 1.4 bar (20 lb./in. 2 ) carbon dioxide or<br />
mixed gas <strong>and</strong> the flow meter coupled with a valve allows an exact pre-set amount of<br />
beer into the keg. A sensor in the return gas pipe keeps the counter pressure gas, <strong>and</strong> so<br />
the fill rate, constant.<br />
Filling by weight can be achieved using load cells incorporated into the line. Many keg<br />
lines incorporate a check-weigher to determine satisfactory operation of the volumetric<br />
system. A weighing platform fits within the conveyor <strong>and</strong> under-weight kegs are rejected,<br />
usually by a pneumatic ram onto a reject line.<br />
21.5.5 Keg capping <strong>and</strong> labelling<br />
A plastic cap is applied to the Barnes neck to protect the filling <strong>and</strong> dispense valve from<br />
dirt <strong>and</strong> to deter tampering. A number of designs are available. Some caps are formed in<br />
the applicator machine whilst other cap types are supplied with an appropriate logo<br />
affixed by the manufacturer. Automatic machines must be capable of 125% of the rate of<br />
the filler so as not to be rate limiting in the overall process. Some caps attach by shrinking<br />
<strong>and</strong> some by clipping to the neck of the keg. The most important aspect is the degree of<br />
security it provides against interference.<br />
Labels on kegs are important in providing the customer with information as to the<br />
beer quality in the keg <strong>and</strong> its `best-before' date, <strong>and</strong> the brewer <strong>and</strong> the customer<br />
with complete traceability of the container. Barcodes applied to paper labels are<br />
essential to contain this information. Machines are now available which, when<br />
programmed, print <strong>and</strong> subsequently apply the labels to the containers. Quality of<br />
printing <strong>and</strong> application is now high. (International Bottler <strong>and</strong> Packer, 2000).<br />
Barcodes can contain the sequential number of the container in the packaging run as<br />
well as the quality <strong>and</strong> best-before date. These barcodes can be scanned at various<br />
points in the supply chain, e.g., at despatch, on the delivery vehicle, at the selling<br />
point <strong>and</strong> on return to the brewery. These labels must be removed prior to reuse of the<br />
keg. This means that the labels can be damaged <strong>and</strong> made unreadable during the<br />
supply chain history of the container.<br />
Containers frequently disappear <strong>and</strong> the paper label will not protect against this.<br />
Brewers have sought more innovative solutions to the problems of container management<br />
<strong>and</strong> traceability. In some systems the entire population of kegs is h<strong>and</strong>ed over to a third<br />
party who, for a fee, manages the population, thus reducing the capital employed by the<br />
brewer (Nelson, 2000b). Radio frequency identification tags can be embedded into the<br />
keg, which identify the keg as a unique container throughout its whole life history. This<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
could be an extremely valuable system if it could be coordinated on anational (or even<br />
international!) scale between competing brewers. Progress on this in the UK has been<br />
slow. Probably further developments will occur in ways of tracking containers, because<br />
of the desire for complete traceability for quality assurance <strong>and</strong> because of the high cost<br />
of containers (approx £40/container).<br />
21.5.6 Smooth flow ale in kegs<br />
As stated earlier, beer for kegging is normally conditioned prior to pasteurization to a<br />
carbon dioxide content of 1.5 to 2.5 vol/vol. Pasteurization <strong>and</strong> filling must then be<br />
managed to maintain this carbon dioxide concentration. In an analogous way to the<br />
development of the widget in canning (Section 21.4.5) there is adesire in the UK <strong>and</strong><br />
Irel<strong>and</strong> to produce keg beers with the drinking characteristics of cask ales. This dem<strong>and</strong>s<br />
acarbon dioxide content of around 1.0 vol/vol <strong>and</strong> athick creamy head. The use of<br />
nitrogen gas provides asolution (Carrol, 1979). The problem with keg beer is getting the<br />
nitrogen intothe beer <strong>and</strong>keeping itin solutionthrough packaging<strong>and</strong>up to thepointof<br />
dispense. Nitrogen can be introduced into aroused bright beer tank at the inlet through a<br />
sinter. The tank will be top pressured with nitrogen <strong>and</strong> considerable manual<br />
involvement will be needed to achieve asatisfactory result (Fitch, 1997). The carbon<br />
dioxide content of the beer must be known <strong>and</strong>, since the nitrogen is being added before<br />
the flash pasteurizer, the pasteurizer must have the pressure capability to keep nitrogen<br />
gas in solution as well.<br />
An improvement is to use amass flow system. Ameasurement is made of the initial<br />
carbon dioxide <strong>and</strong> nitrogen contents of the beer (which should be below specification)<br />
<strong>and</strong> both gases are added. Turbulent flow in the pasteurizer ensures good mixing <strong>and</strong> the<br />
boost pump of the pasteurizer system provides sufficient pressure to prevent gas breakout.<br />
This system dem<strong>and</strong>s very careful control of the operating pressures to achieve the<br />
desired level of nitrogen consistently. In the `Nitroset' system (Lindsay et al., 1995)<br />
nitrogen is injected directly into the beer after the pasteurizer. Aboost pump can achieve<br />
15 bar (220 lb./in. 2 )pressure to keep the gas in solution (Fig. 21.17). Accelerator mixers<br />
create further turbulence to aid solution. Variable flow rates of between 150 <strong>and</strong> 450hl/h<br />
are achievable. Nitrogen concentration can be determined pre- <strong>and</strong> post-addition with<br />
`Orbisphere' thermal conductivity nitrogen monitors.<br />
Further techniques have been described using hydrophobic membranes (Gill <strong>and</strong><br />
Meneer, 1997). The membrane is in the form of ahollow fibre, which allows gas to<br />
diffuse into or out of aliquid without the need for intimate mixing. Each membrane<br />
comprises abundle of fibres with the gas on the inside <strong>and</strong> the liquid on the outside.<br />
Gases can be exchanged into or out of the liquid by varying the partial pressure <strong>and</strong> gas<br />
composition on the inside of the fibres. Pilot scale work has been carried out<strong>and</strong> scale up<br />
to 500hl/hour has been claimed to be feasible. This technique could be valuable for both<br />
the addition of nitrogen <strong>and</strong> the removal of carbon dioxide, particularly if this could be<br />
achieved by asingle pass through the gas exchanger. At present direct injection systems<br />
are the most widely used.<br />
Nitrogenated beer has both carbon dioxide <strong>and</strong> nitrogen in solution <strong>and</strong> these gases<br />
must be kept in solution at the appropriate concentrations up to the point of dispense of<br />
thebeer(Chapter23).Dalton'sLawdem<strong>and</strong>sthatthesametwogasesmustbeinthehead<br />
space of the tank or the keg in the same balanced proportions. The beer can therefore be<br />
filled into the keg using the mixed gas proportion of dispense or, for very low carbon<br />
dioxide beers (< 1 volume), 100% nitrogen can be used.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Press.<br />
reg.<br />
Coalescing<br />
filter to<br />
dry CO 2<br />
Beer supply<br />
Press.<br />
switch<br />
CO 2<br />
purge<br />
N 2<br />
monitor<br />
Infeed<br />
N 2 sensor<br />
Pasteurizer<br />
system<br />
Beer pressure<br />
boost pump<br />
FT<br />
SC<br />
PT<br />
Needle<br />
control<br />
valve<br />
Drain<br />
Flow<br />
indicator<br />
NRV<br />
Gas<br />
filter<br />
Filter<br />
NRV<br />
Press.<br />
switch Press.<br />
Coalescing<br />
filter to<br />
dry N 2<br />
Mass flow<br />
controller<br />
reg.<br />
15 bar<br />
N 2 supply<br />
3 bar<br />
steam<br />
PLC<br />
PI PI<br />
Accelerator – mixers<br />
Press.<br />
reg.<br />
Coalescing<br />
filter to<br />
dry CO2 Press.<br />
switch<br />
CO 2<br />
purge<br />
To SBT<br />
N 2<br />
monitor<br />
Outfeed<br />
N 2 sensor<br />
Fig. 21.17 Flowsheet for the `Nitroset' beer nitrogenation system: FT flow transmitter; SC speed controller; PT pressure transmitter; PI pressure indicator;<br />
NRV non-return valve; PLC programmable logic controller; SBT sterile beer tank (Lindsay et al., 1995).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Atypical `smooth-flow' English ale would have adissolved gas analysis of:<br />
· nitrogen (mg/l) 35 (Range 32 to 38)<br />
· carbon dioxide (vol/vol) 1.1 (Range 1.0 to 1.2).<br />
In the UK virtually the whole of the keg ale trade is in high nitrogen, smooth flow beer.<br />
There is virtually no interest in this type of beer in other parts of the world. There have<br />
been lagers packaged at high nitrogen levels in the UK. These have not been entirely<br />
successful because of the need to maintain asignificant carbon dioxide content (say 2.2<br />
vols) to achieve the sparkle typical of good lager. The nitrogen `softens' the flavour so<br />
that the pleasure of drinking asharp, effervescent lager is lost.<br />
21.6 Cask beer<br />
The final type of packaging to consider is the packaging of naturally conditioned beer<br />
intocasks.Thisbeerisnotfiltered<strong>and</strong>stabilizedinthebrewery.Itcontainsliveyeast<strong>and</strong><br />
depends on asecondary fermentation in the cask to provide condition (carbon dioxide) to<br />
thebeerbefore itisdrunk.Thebeerisnotservedunder pressure<strong>and</strong>isdispensedbyh<strong>and</strong><br />
pumps (Chapter 23). This type of beer is produced in large volumes only in the UK. Of<br />
the 59 million hl of beer produced in the UK in 1998, 38 million hl were sold in draught<br />
form <strong>and</strong> of this volume 11%, about 4.2 million hl was cask beer. Originally beer was<br />
filled (racked) into casks made of wood. These are now rare <strong>and</strong> almost all casks are<br />
stainless steel or aluminium.<br />
Cask beers tend to be rich in flavour <strong>and</strong> aroma, particularly those flavours<br />
associated with hops <strong>and</strong> this makes them unique <strong>and</strong> much sought after by some<br />
drinkers. The advantages to the brewer of producing cask beer relate to the relatively<br />
low cost of the equipment needed <strong>and</strong> the low energy input. Disadvantages include the<br />
inherent variability of the product, its proneness to infection if consumed slowly <strong>and</strong> the<br />
skill required of the publican to dispense it properly (Chapter 23). Some brewery<br />
marketing departments do not like these attributes of cask beer <strong>and</strong> are more<br />
comfortable with the predictable behaviour of keg, bottled <strong>and</strong> canned beer. As a<br />
consequence most ale br<strong>and</strong>s in the UK are sold in keg as well as cask form <strong>and</strong> for<br />
some br<strong>and</strong>s the relative volumes are much more in keg. This trend has been resisted by<br />
a consumer organisation called CAMRA ± the Campaign for Real Ale. This is a<br />
powerful pressure group <strong>and</strong> has done much to persuade brewers to continue to produce<br />
high-quality cask beers. Most brewers seek to develop good relationships with their<br />
local CAMRA branch. CAMRA operates in all areas of the UK.<br />
The issues involved in producing sound cask beer revolve around the h<strong>and</strong>ling of the<br />
casks <strong>and</strong> the h<strong>and</strong>ling of the beer.<br />
21.6.1 The cask<br />
Casks are not pressurized containers <strong>and</strong> are simpler in construction than kegs (Fig.<br />
21.18). Casks contain ahole on the top of the belly called the shive hole that is protected<br />
by the shive boss. Through this hole the cask is filled <strong>and</strong> then closed with the shive,<br />
which can be made of wood or plastic. A second hole on the end of the cask is stoppered<br />
with a plug called a keystone prior to filling. This plug is usually made of wood <strong>and</strong><br />
through this plug is driven, where the beer is to be served, the tap from which the beer<br />
leaves the cask to be drunk (Chapter 23). Casks are usually of 9, 18, or 36 imp. gal.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Protective<br />
bottom chime<br />
∅ 470<br />
Shive boss<br />
Protective<br />
top chime<br />
Keystone<br />
boss<br />
Fig. 21.18 Vertical section of 18 imp. gallon beer cask, height, 520mm; diameter, 470mm (by<br />
courtesy of Alumasc Ltd).<br />
capacity. In the 1970s much beer was sold in hogsheads (54 imp. gal.) but these are now<br />
very rarely seen.<br />
21.6.2 H<strong>and</strong>ling casks<br />
Equipment for h<strong>and</strong>ling empty <strong>and</strong> full casks <strong>and</strong> removing them <strong>and</strong> stacking them on<br />
pallets or spacer boards is identical to the equipment used for kegs. In small breweries<br />
(
Stop/feed<br />
Lifter<br />
Receiving<br />
Bung find <strong>and</strong><br />
pre-rinse<br />
Pre-rinse<br />
To<br />
drain<br />
Ext-clean <strong>and</strong><br />
bung find<br />
HL/det.<br />
HL<br />
21.6.3 Preparing beer for cask filling<br />
Cask beer contains live yeast <strong>and</strong> an important aspect of preparing beer for the cask is<br />
controlling the yeast count at racking. Beer can be racked directly from the fermenting<br />
vessel after skimming <strong>and</strong> an appropriate settling time, but with this arrangement it is<br />
difficult to get aconsistent control of yeast count. After fermentation beer is usually run<br />
into racking tanks or backs. In these tanks the beer is given atime to settle, 16±48h,<br />
depending on the state of the beer <strong>and</strong> dem<strong>and</strong> from trade. The yeast will partly settle.<br />
The objective is to achieve ayeast count at rack of around 1million cells/ml of beer. The<br />
range of yeast count at which cask beer can be successfully packaged is from 0.25 to 4<br />
million cells/ml, but nearer to 1million is to be preferred. Too much yeast suspended in<br />
the beer will result in aviolent secondary fermentation <strong>and</strong> when the casks are vented<br />
prior to sale beer <strong>and</strong> foam will gush from the cask <strong>and</strong> will be lost (Chapter 23). The<br />
remaining beer will be dull <strong>and</strong> lifeless <strong>and</strong> will have little to no head retention or foam<br />
character. If too little yeast is present the secondary fermentation will be too slow <strong>and</strong><br />
there will be insufficient carbon dioxide inthe beer at dispense with the consequence ofa<br />
flat, lifeless beer.<br />
Settling controls yeast count but to aid this process, finings are used (Chapter 15).<br />
Isinglass finings are added at the rate of 1 to 4 pints/imp. brl (0.36±1.44 l/hl). These<br />
finings can be added in the racking tank or at any point up to when the beer is dispensed.<br />
The usual point of addition is at rack with perhaps a prior addition in the racking tank. In<br />
any event the beer will require from 12±48 h <strong>and</strong> possibly up to 72 h to fine <strong>and</strong> settle<br />
before it is sold. The fining of cask beer is one of the most difficult of all brewery<br />
operations to control consistently. Often brewers experience periods of poor fining which<br />
are difficult to explain. Isinglass finings bear a positive charge because of the rich<br />
collagen content <strong>and</strong> interact with the negative charge on the yeast cell wall. In most<br />
circumstances this interaction is sufficient to achieve effective clarity.<br />
Some beers, sometimes will not fine with isinglass alone. The yeast may have a too<br />
low negative charge or the concentration may be too high (say > 2 million cells/ml), or<br />
there may be too high a concentration of positively charged colloids in the beer. In this<br />
HL<br />
To<br />
To reclaim<br />
reclaim<br />
or<br />
re-circulation<br />
Steam<br />
To<br />
drain<br />
Delivery<br />
Lowerer<br />
Rolling rail<br />
discharge<br />
Fig. 21.19 Internal cask washing equipment: nine station chain machine with moving centre<br />
beam; HL, hot liquor (water); det, detergent applied to outside of cask (not used in all installations)<br />
(by courtesy of Porter Lancastrian Ltd.).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
situation auxiliary finings derived from alginates, carrageenan or silicic acid, <strong>and</strong> having<br />
a negative charge, can be added to the beer before isinglass finings to precipitate the<br />
positively charged colloids (Vickers <strong>and</strong> Ballard, 1974). An effective method is often to<br />
add the auxiliary finings in the racking tank <strong>and</strong> separate the flocs thus formed in this<br />
vessel <strong>and</strong> then to add the isinglass at the rack of the beer. Priming sugars are also added<br />
to some beers at this stage. These are normally solutions at 1150 ëSacch (37 ëP) <strong>and</strong> are<br />
added at rates of 1 to 5 pints/barrel (0.35±1.75 l/hl). The priming sugar provides a small<br />
quantity of fermentable carbohydrate (often sucrose) to assist the yeast to achieve<br />
effective secondary fermentation in the cask.<br />
The pH value of cask beer is usually in the range 3.90±4.20. Some brewers add<br />
potassium metabisulphite to beer, which has a bacteriostatic action at pH values below<br />
4.20. This can give some protection against infection but is not a substitute for good<br />
<strong>practice</strong> in cask washing <strong>and</strong> filling.<br />
21.6.4 Cask filling<br />
The size of breweries producing cask beer varies considerably from those producing<br />
national br<strong>and</strong>s at over 5,000 imp. brl (8,000 hl)/week to those producing less than 100<br />
imp. brl (160 hl)/week. Further, micro-breweries or pub breweries may produce only<br />
enough beer to be sold in their own premises <strong>and</strong> this may be only 1±10 brl (2±150 hl)/<br />
week. This means that cask filling operations can vary from single head manual fillers to<br />
large, multi-head racking machines.<br />
Cask beer has a shelf-life from packaging to consumption of a maximum of four<br />
weeks. It contains considerable quantities of yeast, which provides protection against<br />
flavour defects which might be attributable to excess oxygen. While the elimination of<br />
oxygen ingress during packaging is not as critical as with the packaging of brewery<br />
conditioned beers, it is minimized. Modern cask racking machines usually incorporate a<br />
stage of counter-pressure with carbon dioxide gas.<br />
After washing, casks are conveyed by roller or flat chain conveyor to the racking<br />
machine. Prior to racking a keystone is driven into the cask. The casks are rotated bellyup<br />
so that the shive hole is vertically uppermost. The cask is located beneath the filling<br />
tube at a `head' on the racker (Fig. 21.20). There may be up to eight heads on the racker<br />
<strong>and</strong> therefore up to eight casks can be filled simultaneously. The filling tube will<br />
Fob return pipe<br />
Valve<br />
Seal<br />
Sight glass<br />
Beer<br />
Detergent pipe<br />
Fob return vessel<br />
Beer return pipe valve<br />
Beer main<br />
Fig. 21.20 Traditional cask ale racking back (Hough et al., 1982).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
compriseatubeforthebeer<strong>and</strong>atubeforcarbondioxidegas<strong>and</strong>thisisloweredontothe<br />
shive hole to make contact with the cask, which is so sealed from outside air. The cask is<br />
first counter-pressured with carbon dioxide, the bottom valve is opened <strong>and</strong> the beer then<br />
flows at atmospheric pressure into the cask. Air/carbon dioxide in the cask flows through<br />
the return air-pipe, which usually contains asight glass. The cask is deemed to be full<br />
when beer can be seen in the return air-pipe.<br />
Fillers controlled by avolumetric meter are also available. The beer supply is then<br />
shutoff<strong>and</strong>thefillingheadisraised.Beer inthereturnair-pipe isevacuatedtoafobtank<br />
<strong>and</strong>canbeaddedtoafollowingcask.Thisrequiresscrupulousattentiontothecleanliness<br />
ofthemains,asthisprocedureisafrequentsourceofinfectiontothecaskbeer.However,<br />
if this beer is not returned losses will be high. After raising the filling tube the cask is<br />
closedbymanuallydrivingashiveintotheboss.Isinglassfinings(1±4pints/imp.brl)can<br />
also be added at this stage when the beer is said to have been `fined at rack'. This is now<br />
usually the case in the UK because tax (duty) on the beer is paid on the volume <strong>and</strong><br />
strength of the beer leaving the brewery <strong>and</strong> this cannot be controlled if the cask is<br />
broached for fining in adepot or at the point of sale.<br />
Cask beers are renowned for hop character <strong>and</strong> this derives from the particular hops<br />
used in the copper <strong>and</strong> also from the <strong>practice</strong> of dry-hopping beers on the racker so that<br />
hop aroma can develop in the cask in its period of storage prior to dispense. Dry hops are<br />
usually added inthe form of pressed pelletsat rates of 0.5±6ounces (14±84g)/barrel.The<br />
pellets are formed from whole cone hops <strong>and</strong> are lightly pressed to avoid rupture of<br />
lupulin gl<strong>and</strong>s (see also Chapters 7<strong>and</strong> 8). They are supplied in weights of Ü, <strong>and</strong> Ý<br />
ounce (7 <strong>and</strong> 14g). The pellets have ashort shelf-life <strong>and</strong> should be kept at below 15ëC<br />
(60ëF) <strong>and</strong> abatch should be used in four months. The pellets are added by h<strong>and</strong> prior to<br />
closing the cask with the shive. The weight of the pellets <strong>and</strong> hence the essential oil<br />
content is very variable. This has led to the development of various extracts in which<br />
more consistent levels of essential oil can be added (Chapter 7). These products are,<br />
however, difficult to h<strong>and</strong>le on the line, are expensive <strong>and</strong> have not enjoyed wide favour<br />
with brewers particularly as the sales of cask beer are in decline. Many different hop<br />
varieties have been used for dry-hopping <strong>and</strong> particular varieties are favoured for<br />
particular beers. These can be as diverse as East Kent Goldings, Fuggles, Wye<br />
Northdown <strong>and</strong> Styrian Goldings. Brewers should pay particular attention to securing<br />
adequate supplies of hops for several crop years ahead to ensure beer flavour is not<br />
compromised.<br />
After filling the cask is labelled. The comments made on the labelling of beer in kegs<br />
(21.5.5) equally apply to beer in casks. To ensure that the beer is in optimum condition<br />
when drunk requires considerable effort in storage <strong>and</strong> dispense (Chapters 22 <strong>and</strong> 23).<br />
In the UK cask beer remains a favourite choice of experienced drinkers but its wider<br />
appeal is limited by inherent inconsistencies <strong>and</strong> the desire to drink beers at lower<br />
temperatures (< 8 ëC; 46 ëF), which seriously limit the flavour experience of the true<br />
cask product.<br />
21.7 Summary<br />
Packaging is a vital part of brewery operations. The rise in the drinking of beer at home<br />
<strong>and</strong> the influence of retail supermarkets has meant that effective packaging of beer in<br />
bottles <strong>and</strong> cans is essential to catch the eye of the purchaser. Huge amounts of money are<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
spent by global brewers on packaging developments, certainly much more than is spent<br />
on research into processing or raw materials.<br />
In many parts of the world the major package remains the returnable bottle. But local<br />
brewers are in competition with the brewers of international br<strong>and</strong>s putting their<br />
traditional markets under threat by more sophisticated packaging. This has resulted in a<br />
raising of st<strong>and</strong>ards of packaging to preserve the market for local beers.<br />
Even with draught products there is a greater dem<strong>and</strong> by customers for cleaner<br />
containers <strong>and</strong> for labels providing more information about the quality of the beer <strong>and</strong> its<br />
shelf-life. Concurrently the brewer has made efforts to improve the traceability of beer in<br />
kegs or casks, to protect his investment in the container <strong>and</strong> to ensure that the customer is<br />
protected in any cases of poor quality.<br />
<strong>Brewing</strong> becomes increasingly a trans-national business <strong>and</strong> it is likely that as markets<br />
develop the trend will continue towards packaging in non-returnable bottles, probably<br />
made of plastic. This will increase the pressure on finding effective ways of recycling the<br />
empty package.<br />
21.8 References<br />
ATKINSON, B. (1988) J. Inst. <strong>Brewing</strong>, 94, 261.<br />
BEER, C. (1989) Tech. Quart. MBAA. 26, 89.<br />
BLRA (2002) Statistical H<strong>and</strong>book, Brewers' <strong>and</strong> Licensed Retailers' Association, London.<br />
BREWER, A. J. <strong>and</strong> CARTER, A. (2000) Tech. Quart. MBAA. 37, 105.<br />
BROWN, D. (1997) The Brewer, 83, 25.<br />
BUNKER, H. J. (1955) Proc. 5th Congr. Eur. Brew. Conv., Baden-Baden, 330.<br />
BUSH, J.H. (1964) Brewers' Digest, 39, 48.<br />
CARROL, T.C.N. (1979) Tech. Quart. MBAA. 16, 116.<br />
CARTER, A. (1998) Proc. 25th Conv. Inst. Brew. (Asia Pacific Section), Perth, 114.<br />
CARTER, A. (2001) Personal communication.<br />
COLEMAN, M. (1976) Brewers' Guard. 105 (10), 51.<br />
DEL VECCHIO, H. W., DAYHARSH, C. A. <strong>and</strong> BASELT, F. C. (1951) Proc. Ann. Meet. Amer. Soc. <strong>Brewing</strong><br />
Chemists, 45.<br />
DONHAUSER, S. <strong>and</strong> JACOB, F. (1988) Brauwelt, 128, 1452.<br />
DONOVAN, P., CURRIER, R., BLANTON, R. <strong>and</strong> ROSS, J. (1999) Tech. Quart. MBAA. 36, 247.<br />
DUFFY, G. <strong>and</strong> DU TOIT, C. (2000) Tech. Quart. MBAA, 37, 285.<br />
EATON, J. B. (2002) Personal communication.<br />
EVERETT, J. F. (1997) MBAA. Beer Packaging Manual, 167.<br />
FITCH, N. (1997) Ferment, 10 (1), 41.<br />
GILL, C. B. <strong>and</strong> MENEER, I. D. (1997) The Brewer, 83, 77.<br />
HEINS, H. <strong>and</strong> HEUER, J. F. (1997) MBAA. Beer Packaging Manual, 249.<br />
HEUER, J. F. (1997) MBAA. Beer Packaging Manual, 274.<br />
HOUGH, J. S., BRIGGS, D. E., STEVENS, R. <strong>and</strong> YOUNG, T. W. (1982) Malting <strong>and</strong> <strong>Brewing</strong> Science, Chapman<br />
<strong>and</strong> Hall, New York, 721.<br />
HYDE, A. (2000) The Brewer, 86, 248.<br />
International Bottler <strong>and</strong> Packer, (2000) February, 40.<br />
KUNZE, W. (1999) Technology <strong>Brewing</strong> <strong>and</strong> Malting, VLB Berlin, 460.<br />
LANDMAN, B. C. J. (1999) Tech. Quart. MBAA. 36, 329.<br />
LINDSAY, R. F., LARSEN, E. <strong>and</strong> SMITH, I. B. (1995) Proc. 25th Congr. Eur. Brew. Conv. Brussels, 705.<br />
MOLL, M. (1994) Beers <strong>and</strong> Coolers, Intercept, Andover, Hampshire.<br />
MOLL, W. A. (1997) MBAA. Beer Packaging Manual, 83.<br />
NELSON, L. (2000a) Brewers' Guard. 129 (2), 25.<br />
NELSON, L. (2000b) Brewers' Guard. 129 (6), 26.<br />
PALL, D. (1975) Brygmesteren, 32, 197.<br />
PARSONS, M. (2000) The Brewer, 86, 347.<br />
PASTEUR, L. (1876) EÂ tudes sur la bieÁre, Gauthiers Villars, Paris.<br />
RAINBOW, C. (1971) Process Biochemistry, April.<br />
ROUILLARD, C. (1999) Tech. Quart. MBAA. 36, 435.<br />
ROUILLARD, C., <strong>and</strong> HOWELL, M. (1999) Tech. Quart. MBAA. 36, 243.<br />
SCHWARTZ, V. (1997) MBAA. Beer Packaging Manual, 192.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
SCRUGGS, C. E. (1997) MBAA. Beer Packaging Manual, 226.<br />
SPARGO, W. (1997) MBAA. Beer Packaging Manual, 149.<br />
VICKERS, J. C. <strong>and</strong> BALLARD, G. (1974) The Brewer, 60, 19.<br />
WAINWRIGHT, T. (1999) Ferment, 12 (6), 4.<br />
WILLOX, I. C. (1966) J. Inst. <strong>Brewing</strong>, 72, 236.<br />
WILSON, J. R. (1997) MBAA. Beer Packaging Manual, 352.<br />
YEO, A. (2000) Brewers' Guard. 129 (6), 22.<br />
YOKOI, T., SASAKI, T. <strong>and</strong> KURAHSHI, M. (1991) Tech. Quart. MBAA. 28, 12.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
22<br />
Storage <strong>and</strong> distribution<br />
22.1 Introduction<br />
It was commonly the <strong>practice</strong> to deliver beer from breweries directly to customers. This<br />
was the case in the UK <strong>and</strong> parts of Europe from the time of the Industrial Revolution<br />
until well into the 20th century. This was not the case in North America where the<br />
distances to be covered were much greater. The principles of the storage <strong>and</strong> distribution<br />
of beer are governed by the nature of the customer, the distances involved in the delivery<br />
<strong>and</strong> the type of beer. In large breweries the organization of storage <strong>and</strong> distribution is<br />
frequently the responsibility of an entire department, which may even be located away<br />
from the brewery site. As the brewing industry has become more competitive so the<br />
storage <strong>and</strong> distribution of beer has to be carried out with the utmost efficiency <strong>and</strong> at<br />
lowest cost. The likely long-term winners will be the companies who pay most attention<br />
to getting good quality beer to the customer when he wants it, not when it is best for the<br />
brewery to deliver. This has resulted in companies adopting the principles of logistics,<br />
derived from military experience, to ensure optimum deployment of stock <strong>and</strong> vehicle<br />
movement.<br />
22.2 Warehousing<br />
<strong>Brewing</strong> is a capital-intensive business, which involves both fixed assets <strong>and</strong> working<br />
capital. A major part of the working capital is the stock of finished goods held at the<br />
brewery. Beer deteriorates with age <strong>and</strong> so stock levels should be held as low as is<br />
possible consistent with delivery to meet customer requirements. In this way the customer<br />
gets fresh beer <strong>and</strong> the return on capital employed is enhanced.<br />
In most countries, governments take an excise tax on beer so the value of the stock<br />
depends to a large extent on whether it is held as tax paid or tax suspended. In duty<br />
systems where the beer is taxed `at the brewery gate' the beer in the brewery store is held<br />
duty suspended where it might be held duty paid in a store when it has exited the brewery.<br />
It makes financial sense to minimize overall stock <strong>and</strong> maximize duty un-paid stock.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
<strong>Brewing</strong> companieshave become expert atthisbutthedem<strong>and</strong>sofcustomerservicehave<br />
become more important as breweries seek to establish competitive advantage. This has<br />
tended to result in higher stocks of some products being held in order to avoid any<br />
possibility of stock-outs. The dem<strong>and</strong>s on the effective management of the brewery<br />
warehouse have, therefore, increased. Breweries will have awarehouse situated as close<br />
as possible to the end of the packaging lines. Frequently there is aseamless transition<br />
from packaging to warehouse <strong>and</strong> these parts of the brewery may be under the same<br />
management. This avoids disputes as to the rate-limiting steps in the sequence of<br />
operations from packager to customer.<br />
Warehouses vary in complexity in relation to the number <strong>and</strong> types of products<br />
stocked. This can include beer in casks <strong>and</strong> kegs as well as numerous sizes of bottles <strong>and</strong><br />
cans in various package types. This complexity is judged in terms of stock keeping units<br />
(SKUs). Each type of package is one SKU. Beer of the same specification in bottles of<br />
25cl, 33cl, <strong>and</strong> 50cl is 3SKUs. This would further proliferate if those bottles were<br />
packed into boxes containing 15, 18, or 24 bottles, which might be needed for different<br />
customers. Soft drinks might also be h<strong>and</strong>led as well as beer. The number of SKUs can<br />
reach very high levels very quickly. This puts great emphasis on stock control <strong>and</strong> the<br />
organization of production to meet diverse dem<strong>and</strong>s.<br />
Warehouses also have to deal with empty containers returned from trade. These will<br />
include kegs <strong>and</strong> bottles in most cases. These will require sorting <strong>and</strong> cleaning (Chapter<br />
21). Warehousing, therefore, relates to the storage <strong>and</strong> despatch of beer, <strong>and</strong> the<br />
reception, storage <strong>and</strong> issue of packaging materials. These packaging materials can be<br />
quite simple in a warehouse dealing only with large pack beer in cask or keg but<br />
extremely diverse in a warehouse storing small pack beer in bottles or cans in various<br />
types of shrink wrapping <strong>and</strong> boxes, cardboard cases <strong>and</strong> crates.<br />
22.2.1 Principles of warehouse operation<br />
Warehousing requires a large amount of space <strong>and</strong> the multi-h<strong>and</strong>ling of packages. No<br />
matter how low the stock holding, warehousing is expensive, so the area required, though<br />
large, must be kept to a minimum by stacking containers <strong>and</strong> packages on robust pallets<br />
or by using undamaged spacer boards when h<strong>and</strong>ling containers palletless. Space must be<br />
utilized carefully to minimize the extent of mechanical h<strong>and</strong>ling by forklift truck.<br />
Computer systems are widely used to manage the inventory in the warehouse.<br />
Information can be fed to the computer from radio data terminals on the forklift trucks<br />
<strong>and</strong> entries made as beer comes off the packaging line into store <strong>and</strong> leaving the<br />
warehouse floor onto a vehicle.<br />
Stock control<br />
Beer deteriorates with age <strong>and</strong> will have a defined shelf-life that can vary from four<br />
weeks for cask beer to 52 weeks for highly stabilized beer in bottle or can. It is<br />
essential that stock must be rotated on a `first in first out' basis. Most breweries have<br />
strict rules about the age of beer that may be released to a customer, to ensure optimum<br />
quality when the beer is drunk. The exact age on release will depend on how much<br />
stock is held in depots or regional distribution centres. This illustrates just how critical<br />
careful stock control is to the success of the brewery. Beer that goes over age in the<br />
warehouse must either be destroyed (which has implications for excise duty) or<br />
recovered <strong>and</strong> blended back at, say, 5% into mainstream beer. These are both<br />
expensive operations.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Storage conditions<br />
Conditions in the warehouse are critical for beer quality. Filtered beer must be stored<br />
above freezing point <strong>and</strong> at < 22 ëC (< 72 ëF). At low temperatures a chill haze may<br />
develop <strong>and</strong> at temperatures greater than about 22 ëC (72 ëF) flavour stability is<br />
threatened. The storage of cask beer is more critical because of the presence of live yeast<br />
<strong>and</strong> the required secondary fermentation, which continues in the pub cellar. Cask beer is<br />
best stored between 10 <strong>and</strong> 17 ëC (50 <strong>and</strong> 63 ëF). Relative humidity must be kept low by<br />
adequate ventilation. Moist conditions promote the deterioration of secondary packaging<br />
on cases of bottles <strong>and</strong> cans <strong>and</strong> render the beer unsaleable. There must be high st<strong>and</strong>ards<br />
of housekeeping. There should be written hygiene procedures that must be adhered to. In<br />
this way pests such as mice, pigeons <strong>and</strong> cockroaches will be deterred.<br />
Record keeping<br />
In most countries the tax (duty) on beer is paid when the beer leaves the warehouse. It is a<br />
legal requirement to maintain records of the volume <strong>and</strong> strength of the beer being made<br />
<strong>and</strong> subsequently sold. These records must be available for inspection <strong>and</strong> audit at any<br />
time. Governments also usually have legal requirements relating to the volume of beer<br />
sold in packages, which may be related to minimum or average contents. This again<br />
requires record keeping.<br />
Product traceability in trade is now important. Customers are more dem<strong>and</strong>ing of<br />
product quality <strong>and</strong> are frequently encouraged to voice their opinion of the beer by<br />
telephoning help lines. It follows that there must be very sound systems in place for<br />
labelling containers <strong>and</strong> subsequently tracing the beer's history in the brewery, the<br />
warehouse <strong>and</strong> in the trade.<br />
22.2.2 Safety in the warehouse<br />
The warehouse is one of the most dangerous places in the brewery, where frequently the<br />
highest numbers of accidents occur. This is a result of the juxtaposition of forklift trucks <strong>and</strong><br />
the manual h<strong>and</strong>ling needed to get the beer in the right place. In the UK the highest incidence<br />
of `lost-time accidents' is in warehouse <strong>and</strong> distribution work (BLRA, 1987±1995). In the<br />
period 1987±1991 there occurred an average of 1,000 accidents each year resulting in lost<br />
time of more than three days. Of these over half were associated with h<strong>and</strong>ling, lifting or<br />
carrying beer in its various containers. This resulted in the introduction of `Manual H<strong>and</strong>ling<br />
Regulations' in 1993. There are similar regulations in place in other countries. The basic<br />
principle of these regulations is to control risk so far as is reasonably possible:<br />
· avoid hazardous manual h<strong>and</strong>ling where possible<br />
· assess any hazardous operations that cannot be avoided<br />
· remove or reduce the risk of injury so far as is reasonably practicable using the<br />
assessment as a basis for action.<br />
These regulations place a duty on employers <strong>and</strong> employees with the objective of<br />
reducing the number of accidents <strong>and</strong> creating a safer workplace. The employer must<br />
have a safe system of work, which is written down, monitored, audited <strong>and</strong> improved.<br />
Training discharges the responsibility of employers for much of this. The training must be<br />
ongoing <strong>and</strong> this can place a considerable strain on resources of both trainer <strong>and</strong><br />
employee. The rewards are, of course, considerable as the `lost time' of the accidents can<br />
be reduced with big savings in cost <strong>and</strong> improvements in customer service <strong>and</strong> greater<br />
well-being of employees.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
22.3 Distribution<br />
Distribution is one activity that has been subjected to intense scrutiny <strong>and</strong> change in<br />
recent years. This is particularly the case in countries with mature beer markets in Europe<br />
<strong>and</strong> North America. In these areas beer consumption in the early 21st century is falling or<br />
at best is static. Competition between companies is intense <strong>and</strong> all companies seek to call<br />
themselves to the attention of the customer to gain a competitive advantage, however<br />
small. Product quality is of paramount importance. It is simply not possible to sell beer<br />
that is not of the highest st<strong>and</strong>ard. It is very difficult, therefore, to gain advantage on the<br />
basis of product quality. There will be differences in beers between brewers but the<br />
quality will be very similar. Likewise, low-cost production has become almost a given<br />
requirement for the national <strong>and</strong> international brewer. The definition of low cost will vary<br />
from country to country but competitors know the targets in any particular country <strong>and</strong><br />
will measure their performance against those targets by benchmarking to ensure that costs<br />
are under control. It is therefore very difficult to gain advantage on the basis of the cost of<br />
production.<br />
The brewing industry was slow to underst<strong>and</strong> the significance of customer service. In<br />
the UK this was certainly a result of the ownership of the retail outlets for beer. These<br />
outlets were, in the main, in the h<strong>and</strong>s of the brewers of the beer until the report of the<br />
Monopolies <strong>and</strong> Mergers Commission into restrictive <strong>practice</strong> in the brewing industry<br />
was published in 1989. Before this time there was little incentive to devote too much time<br />
to the needs of the customer when that customer was a public house owned by the<br />
brewery. Everything changed when restrictions were placed on the number of retail<br />
outlets that could be owned by beer producers. Competition intensified overnight.<br />
Different pressures applied in other countries of the developed world. One of the main<br />
factors was a major increase in interest in `healthy life-styles' that occurred in the USA<br />
<strong>and</strong> Europe from the early 1980s. Beer sales fell <strong>and</strong> brewers realized they were not only<br />
in competition with each other but with manufacturers of soft drinks. A competitive<br />
advantage was needed <strong>and</strong> brewers looked to customer service to provide it. Much effort<br />
was therefore put into distribution <strong>and</strong> to underst<strong>and</strong>ing the needs of the customer. In<br />
countries where beer consumption is still growing rapidly, such as China, <strong>and</strong> parts of<br />
South-East Asia <strong>and</strong> parts of South America, distribution is still important but it is<br />
unlikely to be the main agent of competitive advantage.<br />
22.3.1 Logistics<br />
Beer supply in breweries was formerly in the h<strong>and</strong>s of a transport department that then<br />
became known as distribution. The controlling interest was frequently the maintenance of<br />
the vehicles rather than the requirements of the customer. Management was mainly<br />
concerned with load planning, i.e., ensuring the optimum weight of beer on the vehicle<br />
<strong>and</strong> ensuring the most efficient route to conserve fuel. The use of logistics is concerned<br />
with optimum deployment of stock <strong>and</strong> vehicle movement. This implies the integration of<br />
planning <strong>and</strong> distribution. The driving force can now be the customer because the<br />
optimization can start with the customer needs <strong>and</strong> stock deployment can be built around<br />
this. Thus the `pull' style of planning develops, starting with the customer requirement<br />
<strong>and</strong> finishing with scheduling the number of brews per week in the brewhouse.<br />
There has generally been a big increase in the dem<strong>and</strong> for small pack beer for<br />
consumption at home (BLRA, 1999), usually bought through supermarkets. This has<br />
introduced a new customer service dimension <strong>and</strong> one that has brought the producer<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
much closer to the consumer. To meet this diverse dem<strong>and</strong> regional distribution centres<br />
away from breweries have been set up in the UK <strong>and</strong> have developed in other parts of<br />
Europe. Regional distribution was always a feature of North American distribution<br />
because of the great distances involved <strong>and</strong> customer service was part of the North<br />
American brewer's way of life earlier than it was in Europe.<br />
Planning<br />
The core of successful logistics is planning. The requirements of the customer must be<br />
interpreted into plans that can deliver the right amount of stock at the right time <strong>and</strong><br />
efficient productionofthatstockbythebrewery.Theplanisderived fromhistorical sales<br />
data <strong>and</strong> forecasting. The former is amatter of fact <strong>and</strong> the latter is frequently amatter of<br />
conjecture! Software systems are, however, now available that can be used to telling<br />
effect in forecasting customer needs. This is important because for draught beers<br />
customers often run on very low stocks so that their working capital is reduced. The<br />
customer expects the supplier to deliver the beer immediately he requires it <strong>and</strong> in the<br />
appropriate state for dispense. In other words the customer wants the brewer to hold the<br />
stock. Good planning systems can turn this to advantage by using the principle of `just in<br />
time' delivery so that brewery stock is also minimized. `Just in time', must not become<br />
`just too late'! If cask beer is being supplied then this is more difficult because of the<br />
finite time required for adequate secondary fermentation (Chapter 23).<br />
Customer dem<strong>and</strong> is the driving force behind most logistic systems. However, for<br />
national big br<strong>and</strong>s (> 800,000 hl pa, approx. 500,000 imp. brl) a system of `vendor<br />
managed inventory' can be used. Here the supplier controls the stock in the supply chain<br />
<strong>and</strong> delivers to the customer depending on the supplier's view of the customer's<br />
requirements. This requires having software to monitor stock at the customer's premises<br />
<strong>and</strong> delivering when that stock requires replenishment. This means that the beer should<br />
be in the right place to meet dem<strong>and</strong> <strong>and</strong> should guard against the customer who orders<br />
beer when, based on his sales, he does not really need it. Vendor managed inventory can<br />
thus prevent spurious shortages in the supply chain.<br />
Delivery<br />
Traditionally two types of delivery from breweries were recognized, primary <strong>and</strong> retail or<br />
radial. Radial delivery was very common in the UK where pubs were situated close to<br />
breweries <strong>and</strong> a retail fleet of vehicles would be based at the brewery <strong>and</strong> would make<br />
several trips to customers probably five or six days a week returning to the brewery each<br />
time with empty containers. This is still a major feature of regional breweries in Britain<br />
<strong>and</strong> the rest of the world.<br />
Primary delivery constitutes delivery of beer from the brewery to a depot or regional<br />
distribution centre where stock is held. This is usually by heavy articulated vehicles with<br />
gross weights of up to 40 t. Draught beer is most often h<strong>and</strong>led palletless, with groups of<br />
containers separated by plastic spacer boards. This reduces weight <strong>and</strong> allows a greater<br />
payload of beer.<br />
As brewers have sought to differentiate their service to gain commercial advantage,<br />
the composite delivery has become common. In this system the brewer offers to deliver a<br />
composite mixture of beer in all package types <strong>and</strong> soft drinks <strong>and</strong> wine <strong>and</strong> spirits. This<br />
requires having the facility to do this at the brewery or depot. A retail or radial fleet must<br />
operate from the depot to the customer but frequently loads can be `made-up' for the<br />
`primary' vehicle <strong>and</strong> merely transferred to the retail vehicle <strong>and</strong> delivered to the<br />
customer.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Afurther complication is the delivery of beer by road tanker to asecondary location<br />
for further packaging. This might be beer for kegging or bottling or canning. In this<br />
situation the supplying <strong>and</strong> receiving departments must agree on the rules for delivery<br />
<strong>and</strong> acceptance to avoid the costly business of returning unemptied tankers to the source,<br />
which is to nobody's benefit. This is all the more relevant when the supplier <strong>and</strong> the<br />
receiver are part of the same company.<br />
This whole business is about delighting the customer whilst operating at the lowest<br />
cost. Organization has been helped by seeking the registration of the business to quality<br />
st<strong>and</strong>ards such as ISO 9002. These systems do not, of course, result in improved quality<br />
to the customer's benefit per se but they do provide awritten guide to asystem of work<br />
that can be audited <strong>and</strong> subjected to continuous improvement. Another benefit of these<br />
systems resides in the fact that if they are not adhered to the registration can be removed<br />
bytheauditingauthority<strong>and</strong>thatwouldbeaconsiderablesourceofembarrassmenttothe<br />
brewer <strong>and</strong> could put him at severe disadvantage. There is, therefore, every incentive to<br />
maintain <strong>and</strong> improve the system.<br />
22.3.2 Quality assurance<br />
The second major aspect of modern distribution along with logistics is the assurance of<br />
quality. This is made much easier if the beer in the brewery is produced to sound<br />
st<strong>and</strong>ards. Customers rightly expect their beer to be delivered to the highest quality<br />
st<strong>and</strong>ards <strong>and</strong>, irrespective of any nationally approved quality management system,<br />
reserve the right to inspect the brewers' premises at any time. Quality is assured most<br />
effectively by the brewer having in place strict st<strong>and</strong>ards for the release of beer to<br />
delivery vehicles, whether for primary or retail supply. There should be aprocedure in<br />
place the operation of which prevents sub-st<strong>and</strong>ard beer being delivered. This requires<br />
setting specifications for the beer in the warehouse <strong>and</strong> ensuring that the beer meetsthose<br />
specifications before it is delivered. This should avoid the requirement for recall of beer<br />
from the supply chain. No matter how good the system, however, some beer will be<br />
released which will not be of the right quality. It follows that product recall procedures<br />
must also be in place to assure the customer that the brewer is in charge of quality<br />
throughout the supply chain.<br />
There will need to be some tolerance in the analytical values of the beer so as to<br />
prevent unnecessary hold-ups in supply. The system must also allow for `beer on hold'<br />
whilst further analysis is carried out. Amost important aspect of this release of beer to<br />
trade is the flavour of the beer. Aproperly trained panel, as part of the quality assurance<br />
beer release process, must taste all beers. Typical beer release parameters for alager beer<br />
are shown (Table 22.1). Analyses will be determined on samples drawn from each brew.<br />
Beer is likely to be released to primary delivery throughout the day. It is very<br />
important therefore that people on every shift in the warehouse or packaging department<br />
are trained in beer release procedures. They must have an underst<strong>and</strong>ing of the analytical<br />
values in Table 22.1. Of course, the beer should not be packaged if it does not meet the<br />
required specification. The required specification is often achieved by blending the<br />
contents of different bright beer tanks of the same beer quality. The accurate setting of the<br />
blend <strong>and</strong> the computation of the weighted average analysis are critical for success.<br />
It is clear that quality assurance for distribution dem<strong>and</strong>s effective quality assurance<br />
throughout the supply chain in the brewery. This in turn dem<strong>and</strong>s effective<br />
communication from the brewhouse through to fermentation <strong>and</strong> to beer processing<br />
<strong>and</strong> packaging. Beer should not move along the supply chain unless it meets the in-<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table 22.1 Release parameters for a lager beer<br />
Parameter Target value Permitted range<br />
OG (ëSaach) 40 38.5±41.5<br />
PG (ëSaach) 8 6.5±9.5<br />
ABV (%) 4.1 3.9±4.3<br />
pH 4.0 3.8±4.2<br />
Colour (ëEBC) 8.5 7.0±10.0<br />
Bitterness (ëEBU) 19 16.5±21.5<br />
CO2 (vol.) 2.1 2.0±2.2<br />
Dissolved oxygen in BBT (mg/l) 6 5 min.<br />
Haze (ëEBC, Monitek)
23<br />
Beer in the trade<br />
23.1 Introduction<br />
In recent years there have been large increases in the consumption of beer at home in<br />
bottles <strong>and</strong> cans (BLRA, 1999a). This has been particularly the case in Europe but has<br />
also occurred in North America. There has also been an increase in the consumption of<br />
beer in bottles in licensed premises. This has been associated with a greater emphasis on<br />
food being served along with beer. Frequently this beer is drunk directly from the bottle,<br />
which is usually of dark brown or green glass. In this situation the beer cannot be seen,<br />
which denies the drinker one of the pleasures of drinking beer, that of contemplating its<br />
appearance in the glass. Despite these trends the quality of draught beer drunk in the<br />
trade, i.e., in public houses <strong>and</strong> the like, is very important. Quality is influenced not only<br />
by the quality of the beer delivered but also by the conditions of storage in the public<br />
house <strong>and</strong> the conditions of the dispense of the beer. This applies equally to cask beer<br />
where a controlled secondary fermentation takes place in the container <strong>and</strong> to keg beer<br />
dispensed from a sterile container by gas top pressure.<br />
Advertising of national <strong>and</strong> international beers centres on creating an `image' of<br />
br<strong>and</strong> values which must be consistent. Brewers simply cannot leave the control of the<br />
quality of their beer to chance. They must ensure that good <strong>practice</strong> is applied in storage<br />
<strong>and</strong> at dispense, so that when the beer is drunk the intended quality is attained. This can<br />
be done by employing teams of people to advise customers on best <strong>practice</strong> <strong>and</strong> to<br />
ensure it is effected or to contract out this operation to a specialist company. Both<br />
systems are satisfactory. The owner of a strong br<strong>and</strong> can insist on the quality st<strong>and</strong>ards<br />
of storage <strong>and</strong> dispense being a prerequisite to allowing the customer to sell the<br />
product.<br />
Beer is consumed in a wide variety of premises. These can vary from large city<br />
centre public houses <strong>and</strong> bars with extensive cellar facilities to beach bars where beer is<br />
served in ambient temperatures of over 30 ëC (86 ëF) with a simple on-line cooler. It<br />
follows that there must be some basic principles of management of beer in the trade if<br />
the brewer is going to satisfy the maximum number of customers in this variety of<br />
outlets.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
23.2 History<br />
Home brewing diversified into acraft industry to serve local communities. At this time<br />
all beer was dispensed on draught. Frequently the beer would be collected from the<br />
brewery in pots or jugs <strong>and</strong> then dispensed at home. Taxes also had an influence. In the<br />
UK there was atax on glass in the 19th century, which was removed at the start of the<br />
20th century, <strong>and</strong> this led to brewers examining the potential of bottling beer. In the UK<br />
thetraditionalcontainerforbeerwasthecask,originallymadefromwoodbycoopers<strong>and</strong><br />
now usually made of stainless steel or aluminium (Chapter 21). There is still adem<strong>and</strong><br />
for cask beer but by 1998 this type of beer represented about 10% of total draught beer<br />
sales in the UK (BLRA, 1999b). In spite of intense interest in this type of beer by microbrewers<br />
all over the world cask beer is much less significant in other countries <strong>and</strong><br />
certainly so in the countries of largest production: USA, China <strong>and</strong> Germany. In these<br />
countries, <strong>and</strong> in most others, the most important type of draught beer is that dispensed<br />
from apressurized container by gas top pressure in astabilized <strong>and</strong> filtered form. This<br />
applies to both lager <strong>and</strong> ale.<br />
The development of keg beer in the UK was stimulated by the presence of US airmen<br />
in the Second World War. Apparently they did not take to traditional British cask beer.<br />
This prompted development work by the brewer Greens of Luton who refined the process<br />
of putting sediment free, carbonated beer into metal containers <strong>and</strong> so keg beer was<br />
created (Bamforth, 1998).<br />
23.3 Beer cellars<br />
Traditionally beer was served from cellars. In the main this allowed for better<br />
temperature control <strong>and</strong> provided a space in which the beer containers could be organized<br />
<strong>and</strong> proper stock control carried out. For bars serving large volumes of beer in a week,<br />
say > 15 hl (9 imp. brl), an organized space is essential or it will be impossible to manage<br />
stock rotation <strong>and</strong> ensure that the beer is served on a `first in first out' basis. The `cellar'<br />
will not be below ground level in all bars but the principles of cellar design will still<br />
apply.<br />
In most countries beer is classified as a foodstuff (Hunter, 1993) <strong>and</strong> so is subject to<br />
some food hygiene regulations. This means regular inspection by agents of local or<br />
national government. It follows that companies responsible for the management of the bar<br />
in which the beer is sold must have st<strong>and</strong>ards for the beer cellar which must be written<br />
down, continually assessed, <strong>and</strong> improved. This will be to the ultimate advantage of the<br />
customer <strong>and</strong> should be part of the way the brewer seeks to gain competitive advantage.<br />
23.3.1 Hygiene<br />
A beer cellar should be kept clean, <strong>and</strong> cleaning is greatly aided if the original design is<br />
sound. It is therefore surprising how infrequently very clean cellars are encountered.<br />
Floor surfaces should be hard, impervious, <strong>and</strong> smooth <strong>and</strong> should slope to a drain or<br />
sump. This makes regular hosing of the floor easy to manage. The sump should be<br />
cleaned weekly. Walls <strong>and</strong> the ceiling should have a smooth finish <strong>and</strong> can be tiled or<br />
treated with an anti-fungicidal paint. This allows for frequent washing <strong>and</strong> disinfection<br />
<strong>and</strong> the inhibition of mould growth. All equipment should be stored carefully <strong>and</strong> in a<br />
clean place away from the main area of beer dispense. Beer cellars should not be used for<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
storing other foodstuffs. Bacteria <strong>and</strong> wild yeast can be present on foods such as cheese<br />
<strong>and</strong> these can infect <strong>and</strong> will affect particularly cask beer.<br />
23.3.2 Temperature<br />
Temperature control in the brewery warehouse has been discussed (Chapter 22).<br />
Temperature control in the beer cellar is equally important if the beer is going to be at its<br />
bestwhenserved.Kegbeerwillalmostcertainlybepassedthroughabeercoolerenroute<br />
to the glass. Nevertheless, the workload of that cooler will be reduced if the temperature<br />
of the cellar is controlled. Beers will deteriorate more slowly in acool cellar.<br />
Cask beer requires more defined temperature control to allow effective secondary<br />
fermentation to take place <strong>and</strong> to prevent the development of flavour defects <strong>and</strong> to guard<br />
against possible infection as yeast activity subsides. The ideal cellar temperature to<br />
satisfy all these requirements is between 10 <strong>and</strong> 17 ëC (50 <strong>and</strong> 63 ëF). Temperatures were<br />
originally controlled by ventilation alone but now a cellar refrigeration unit is used. This<br />
should be fan assisted <strong>and</strong> must be properly sized for the cellar volume. Heat generating<br />
units of any sort (e.g. ice makers) should not be located in the cellar so as not to compete<br />
with the temperature control unit.<br />
Ventilation of the cellar or storage area is important in its own right. Air, which is not<br />
changed, becomes stale, humid <strong>and</strong> musty <strong>and</strong> can promote the growth of moulds. There<br />
is also the safety hazard for the dispense of keg beers associated with the use of carbon<br />
dioxide <strong>and</strong> nitrogen. A rise in concentration of these gases can be fatal for persons<br />
working in cellars <strong>and</strong> they must be removed by ventilation <strong>and</strong> there must be training for<br />
staff to cover all aspects of the dangers of these gases.<br />
23.3.3 Lighting<br />
Cellars must be adequately lit <strong>and</strong> the level of lighting required is frequently laid down by<br />
statute. The best lighting is provided by fluorescent tubes with splash-proof covers<br />
(Hunter, 1993). In larger cellars emergency lighting is now often installed.<br />
23.4 Beer dispense<br />
Beers are judged visually as well as by flavour <strong>and</strong> aroma. How a beer appears in the<br />
trade is greatly influenced by how it is dispensed. This applies to both cask <strong>and</strong> keg beer<br />
<strong>and</strong> it is controlled by both the dispense equipment design <strong>and</strong> how it is used. This can be<br />
subject to historical regional influences. In the north of Engl<strong>and</strong> drinkers generally prefer<br />
cask beer with a tight creamy head. In the south of Engl<strong>and</strong> a loose open head is often<br />
preferred. As brewers seek to differentiate their products <strong>and</strong> gain competitive advantage<br />
greater emphasis has been placed on the quality of dispense. This is a key factor in<br />
defining the br<strong>and</strong> <strong>and</strong> major br<strong>and</strong>s usually have unique dispense equipment <strong>and</strong> defined<br />
conditions of use.<br />
Many of the widely differing beer selling premises are not owned by brewers, indeed,<br />
the beer may be supplied by a general wholesaler. Often staff serving beer are employed<br />
on a temporary basis or may not have received brewery training. It follows that the beer<br />
dispense equipment must be robust <strong>and</strong> tamper-proof if the brewer is to be satisfied with<br />
the presentation of his beer. It is also important that waste is minimized because fierce<br />
competition has led to a general reduction in the price the brewer obtains for his beer. In<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
turn this has led to a dem<strong>and</strong> for a more robust cask beer that is easier to h<strong>and</strong>le in the bar<br />
<strong>and</strong> less prone to wastage <strong>and</strong> dispense problems.<br />
All recent developments in the dispense systems of keg <strong>and</strong> cask beers have sought to<br />
guarantee the quality in all kinds of places selling beer. However, there are aspects of the<br />
equipment that if neglected will lead to poor beer quality in the trade with a consequent<br />
loss of immediate sales <strong>and</strong> an adverse effect on the long-term market share of the br<strong>and</strong>.<br />
23.4.1 Keg beer<br />
Keg beers, ales or lagers, are packaged into sterile metal containers of size 25±100 hl,<br />
although in the UK barrels are still used (36 imp. gal or 163 hl). The beer has a defined<br />
content of carbon dioxide <strong>and</strong> will normally be dispensed by applying a top pressure of<br />
carbon dioxide gas or by using a mixed gas usually containing carbon dioxide <strong>and</strong><br />
nitrogen. This is commonly called the `free-flow' system. Pumps can also be used to<br />
assist the flow. A further refinement of the pumped system is to use a metered dispense<br />
where the beer is dispensed in pulses of a set volume, often 250 ml or, in the UK half an<br />
imperial pint (284 ml). Lagers are usually dispensed in a free-flow system with carbon<br />
dioxide alone whilst ales can be dispensed with carbon dioxide or with a carbon dioxide/<br />
nitrogen mixture. This presents the first challenge in assuring the quality of dispense.<br />
Carbon dioxide<br />
Every keg beer has a CO2 specification. There must be careful adjustment of the pressure<br />
of CO2 entering the keg to ensure perfect dispense of the product. The beer will be<br />
susceptible to picking up carbon dioxide or losing it. In either case dispense problems<br />
will follow <strong>and</strong> poor quality will result. If there is insufficient top pressure gas will break<br />
out of the beer into the head space of the keg <strong>and</strong> into the beer lines up to the dispense<br />
tap; a flat beer will result. If the top pressure is too high then more CO2 will dissolve into<br />
the beer than is breaking out <strong>and</strong> the beer will over carbonate with consequent fobbing<br />
problems at the dispense tap.<br />
Temperature is critical when using 100% carbon dioxide for dispense. If the temperature<br />
of the beer rises 1 ëC (1.8 ëF) then an additional 0.1 bar (1.47 lb./in. 2 ) of pressure is needed.<br />
If the temperature falls by 1 ëC then the pressure needs to be reduced by 0.1 bar.<br />
Temperature fluctuation is a common cause of beer dispense problems <strong>and</strong> one of the most<br />
frequent causes of complaints to the brewery. It is not a brewery problem provided that the<br />
carbon dioxide content of the beer was in specification on delivery. It is easy to eradicate in<br />
the trade by attention to the details of temperature control in the cellar.<br />
Problems also occur with beer dispense in high ambient temperatures (Stanley, 1999).<br />
The temperature of the beer line to the tap is often at least 3 ëC (5 ëF) higher than the cool<br />
cabinet or cellar where the kegs are stored. Carbon dioxide will therefore break out of the<br />
beer in the dispense line <strong>and</strong> excessive fobbing at the tap will be the result. In this<br />
situation the CO2 pressure should be set to be appropriate for the lower temperature. If the<br />
cool cabinet is at 2 ëC (36 ëF) <strong>and</strong> the beer line at 5 ëC (41 ëF) then the top pressure should<br />
be increased from 0.7 bar to 0.9 bar (10 to 13 lb./in. 2 ). This would apply to a beer with,<br />
say, a CO 2 content of 2.6 vol/vol. A further complication is that over a period of a few<br />
days <strong>and</strong> as a consequence of the higher top pressure of the gas the carbonation in the<br />
beer will increase to about 2.9 vol/vol <strong>and</strong> fobbing at the tap will be the result. Pressure<br />
compensations are also required when forcing beer by top pressure along long runs of<br />
horizontal or vertical pipe. For every horizontal metre an additional 0.011 bar (0.16 lb./<br />
in. 2 ) is required <strong>and</strong> for every vertical metre an extra 0.108 bar (1.59 lb./in. 2 ).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Thus it is important to pay attention to the correct dispense pressures for the<br />
appropriate temperatures <strong>and</strong> lengths of installation pipe when using 100% carbon<br />
dioxide for dispense. It is highly desirable to maintain the beer at as constant a<br />
temperature as possible <strong>and</strong> to avoid major temperature differences between the beer in<br />
the store <strong>and</strong> the dispense lines to the tap.<br />
Mixed gases<br />
If the beer temperature cannot be held constant then dispense problems will always occur<br />
when using 100% carbon dioxide as the top pressure gas. Aresolution to this problem is<br />
to introduce another gas to the headspace of the keg thus increasing the system pressure<br />
without the potential for increasing the carbon dioxide content of the beer. This can be<br />
done by using nitrogen <strong>and</strong> by using the correct blend of nitrogen <strong>and</strong> carbon dioxide.<br />
The partial pressure of the CO 2is enough to maintain the required carbonation level of<br />
the beer. Nitrogen is not as soluble as CO2 <strong>and</strong> will not affect the taste of the beer unless<br />
used at excessively high pressures. This system of mixed gas dispense can be used to<br />
dispense avarietyofpale beersofdelicateflavour invarying ambientconditions, without<br />
over carbonation. If the beer temperature cannot be controlled to better than Ô2ëC<br />
( 3.6ëF) then ablend of 80% CO2 <strong>and</strong> 20% nitrogen at 1.03 bar (15 lb./in. 2 )will<br />
eliminate dispense problems on a2.6 vol/vol CO2 beer.<br />
Acommon system for using mixed gas is to use ablender to accurately proportion the<br />
gases from cylinders. These must be very carefully set <strong>and</strong> not tampered with. (The best<br />
systems will not operate when one of the gases has run out!) Pre-mixed gas can also be<br />
used, but this is often much more expensive than blending on site. The blends are<br />
sometimes not precise <strong>and</strong> are often low in carbon dioxide. They must therefore be<br />
checked with agas analyser before use.<br />
Adifferent way of exploiting the use of mixed gas has been developed in the UK.<br />
Small amounts of nitrogen (10±50mg/l) have amajor positive effect on beer foam<br />
stability (Bishop et al., 1975) <strong>and</strong> stability of draught stout is much improved by the<br />
inclusion ofnitrogen in the beer. This principle has been extended to ale. This has largely<br />
been amarket-led initiative to improve foam quality on keg ale <strong>and</strong> so provide aproduct<br />
similar in appearance to cask beer drawn by h<strong>and</strong> pulling. Nitrogen is added to filtered<br />
beer prior to packaging (Chapter 21) usually at rates of 15±40mg/l. Normally this beer<br />
has a carbon dioxide content of around 1.1 vol/vol which is much lower than keg beers<br />
not containing nitrogen (up to 2.6 vol/vol).<br />
Nitrogen-containing draught ales are dispensed with a mixed gas blend of carbon<br />
dioxide <strong>and</strong> nitrogen. The nitrogen now has the role of avoiding the loss of nitrogen in the<br />
beer during its dispense life as well as providing the motive force to drive the beer to the<br />
dispense tap without using excessively high levels of carbon dioxide (Lindsay et al.,<br />
1996). Ale is not drunk in large volumes outside the UK <strong>and</strong> so this technique is of local<br />
interest. The introduction of nitrogen to lager beer has been tried but, because of the<br />
softening of the palate of the beer, has not always been commercially successful.<br />
Dispense problems with nitrogenated lager can be severe because of the much higher<br />
carbon dioxide level in the beer (2.6 vol/vol) which, if lowered, produces an insipid<br />
mouth-feel.<br />
Beer pumps<br />
Beer pumps can also be used to protect against the effects of temperature changes<br />
between the beer store <strong>and</strong> the dispense tap. Mechanical pressure is used to push the beer<br />
to the tap rather than the gas top pressure on the keg. This avoids the use of any gas other<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
than carbon dioxide. Pressure on the keg can be set at 0.8 bar (12lb./in. 2 )<strong>and</strong> can be<br />
easily increased to 1.4 bar (20lb./in. 2 )in the beer line by apump. Systems using pumps<br />
require more maintenance, <strong>and</strong> hence are more costly to operate than systems relying on<br />
gas pressure alone. Adevelopment of the pump particularly used in the UK was the<br />
positive displacement meter. Beer was dispensed at high speed, in fixed pulses usually of<br />
one half imperial pint (284ml). These systems are not being widely used. They are<br />
expensive to maintain. Further, in countries where the head on the beer is ruled as being<br />
part of the beer measure then the meter is difficult to operate to customer satisfaction.<br />
The properly balanced mixed gas system offers the most flexible solution to providing<br />
excellence in dispense at lowest cost without the danger of over carbonation.<br />
Beer lines<br />
Abeer line is the route of supply from the storage container in the cellar to the dispense<br />
tap on the bar. The line itself is usually made of polythene or PVC tubing of appropriate<br />
diameter. Poordesign of the beer line or poor cleaning will adversely affect quality ofthe<br />
dispensed beer. The system should be designed <strong>and</strong> adjusted to give aflow rate of about<br />
12.5sec for 500ml (14 sec/imp. pint) of beer (Heron, 1992). This dem<strong>and</strong>s balancing the<br />
top pressure used against the restriction provided by the line. Although this time period is<br />
a good `rule of thumb' competitive advantage is sometimes sought by brewers by<br />
deviating widely from this norm. This particularly applies to nitrogenated ales where<br />
longer dispense times are dem<strong>and</strong>ed to heighten anticipation of the drink.<br />
Long beer lines (>30m; 98ft.) should be cleaned using an installed system involving<br />
re-circulation of the cleaning fluid. Acid based cleaners should be used frequently to<br />
avoid the accumulation of beer stone. Transparent pipe is best so that accumulation of<br />
soil can be seen easily <strong>and</strong> gas `break out' observed. There have been recent<br />
improvements to line cooling systems. Originally line cooler units were situated below<br />
the bar. Cooling used achilled water bath with the beer passing through in coils of tubing<br />
immersed in the chilled water. Further developments used ice-bank coolers with the beer<br />
subject to rapid cooling by ice to supply outlets with high dem<strong>and</strong>. Modern units have<br />
beer re-circulation pumps to allow variable temperature control on ales <strong>and</strong> lagers.<br />
Amajor developmenthasbeen the use ofremotecoolers, which have the advantage of<br />
removing the heat source from below the bar. They are usually situated immediately<br />
outside the cooled beer cellar. Beer is supplied to the bar through an insulated pipe<br />
containing cooling water known as a`python'. Large systems can accommodate up to 16<br />
separateroutesforbeer,knownasproductcoils(Fig.23.1).Theprimarylagerpythoncan<br />
be held at a lower temperature than the secondary ale python, which can take recirculation<br />
water from the lager line to save energy (Fig. 23.2). Pythons are now<br />
sometimes used with under-counter cooling modules, which can further trim beer<br />
temperature, <strong>and</strong> they will take chilled water from the python.<br />
Manufacturers of keg beer dispense systems usually provide advice on the operation<br />
<strong>and</strong> cleaning of their equipment. Follow that advice! Many problems of beer quality<br />
encountered in bars are simply a result of poor cleaning or of operating systems at the<br />
wrong pressures or with poor temperature control.<br />
23.4.2 Cask beer<br />
The cask beer market in the UK still amounts to the considerable annual volume of about<br />
6 million hl (3.7 m imp. brl). Cask beer requires very different treatment in the trade from<br />
keg beer. The major factor affecting cask beer dispense is that cask beer contains live<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Water bath<br />
temperature<br />
gauge<br />
Control panel<br />
Thermostats<br />
Water pump<br />
Filler cap<br />
Condenser<br />
Agitator<br />
Fan<br />
Pump water return<br />
Compressor<br />
Product coil<br />
Fig 23.1 Large remote beer dispense cooler (Lindsay, 2002).<br />
C<br />
B<br />
A<br />
ALE LAGER<br />
D<br />
F<br />
G<br />
RETURN<br />
FLOW<br />
H<br />
E<br />
Filler cap<br />
Evaporator<br />
Product coil<br />
Compressor<br />
electrics<br />
Fig 23.2 Secondary keg ale python installed with a remote beer cooler <strong>and</strong> lager python. A,<br />
kegs; B, keg connector; C, PVC or polythene tubing; D, ale python; E, lager python; F, ale<br />
python water re-circulation line (return); G, ale python water re-circulation line (flow); H,<br />
python driver; I, remote cooler; J, lager python water re-circulation line (return); K, lager python<br />
water re-circulation line (flow); L, lager product line. The ale python takes re-circulation water<br />
from the lager line (Lindsay, 2002).<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
L J<br />
K<br />
I
yeast to effect the secondary fermentation to give condition to the beer. Cask beer,<br />
therefore, has ashelf-life of only 3±4 weeks <strong>and</strong> everything must be geared to ensuring<br />
that the beer is drunk in perfect condition within this time period.<br />
Delivery of beer<br />
An important point is to check that the beer being delivered is within the age profile laid<br />
down by the brewery so that, given time to settle, it will be dispensed within the shelf-life.<br />
This requires knowledge of the throughput of the bar. Containers should not be leaking <strong>and</strong><br />
the shives <strong>and</strong> keystones (Chapter 21) should be intact <strong>and</strong> clean. The casks should be<br />
located by the draymen in the position where they are to be dispensed. The more they are<br />
moved the less satisfactory will be the fining action on which beer clarification depends.<br />
Stillaging<br />
Casks should be positioned on the level on the stillage (a wooden or metal frame on<br />
which the casks are raised off the floor of the cellar) with the shive uppermost <strong>and</strong> the<br />
keystone centrally located at the bottom. The casks should then be secured with wooden<br />
wedges (scotches). The sediment will now collect in the belly of the cask. Sometimes a<br />
tilting stillage is used where the cask can be progressively tilted forward as the beer is<br />
dispensed. This has to be done with care or the sediment will block the dispense tap.<br />
An alternative system is to stillage vertically with the keystone uppermost (Hunter,<br />
1993). The beer is subsequently dispensed through ahollow rod <strong>and</strong> so the cask must be<br />
tilted by positioning awedge directly under the location of the keystone. The sediment<br />
then collects away from the rod inlet. Purists would argue this is not genuine cask beer!<br />
Pegging (spiling)<br />
The secondary fermentation in the cask produces carbon dioxide, which causes arise in<br />
pressure in the container. This pressure should be released gently or the sediment in the<br />
cask can be excessively disturbed when the cask is tapped for dispense.<br />
The shive should first be scrubbed clean <strong>and</strong> then a soft porous peg is driven carefully<br />
into the depression in the shive (tut). This will allow the beer to `work' gently as the carbon<br />
dioxide escapes through the porous peg. This operation should be carried out between three<br />
<strong>and</strong> six hours from delivery. The soft pegs should normally be changed twice a day. The<br />
beer may continue to give off gas for 12±24 hours. When this working has stopped the soft<br />
peg should be removed <strong>and</strong> replaced with a non-porous, hard peg. The hard pegs should be<br />
eased daily to prevent further pressure build up. Best <strong>practice</strong> dem<strong>and</strong>s the insertion of a<br />
soft peg whilst the beer is being dispensed as this acts as an air filter. Between dispense<br />
sessions a hard peg should be inserted to maintain beer condition.<br />
Tapping<br />
The keystone must first be scrubbed clean <strong>and</strong> then a clean tap should be smartly driven into<br />
the already pegged cask with a single hammer blow. A tap should never be inserted into an<br />
un-pegged cask. The beer should be tapped about 12±24 hours before it is required for sale.<br />
This can, of course, vary but most cask beers will have dropped bright (yeast <strong>and</strong> sediment<br />
will have settled) <strong>and</strong> have stopped working in 12±48 hours from delivery although on<br />
occasions 72 hours may be needed. Before connection to the dispense line about 250 ml (<br />
Ý imp. pint) of beer should be drawn off through the tap to clear any sediment around the<br />
tap. This beer should be discarded. A further 250 ml ( Ý imp. pint) should then be drawn<br />
<strong>and</strong> checked for clarity, aroma, <strong>and</strong> taste. If the beer is found to be true to type the container<br />
can be connected to the dispense line so the beer can be sold.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Tilting<br />
Stillaged casks should be tilted forward when about one-third of the contents of the cask<br />
have been sold <strong>and</strong> not before. Afall of about 7cm (2.75in.) is needed. This allows the<br />
maximum yield of beer from the cask. The task should be done carefully so as not to<br />
disturb sediment.<br />
Dispense<br />
Originally beer was dispensed into ajug through the tap from the stillaged cask at the<br />
back of the bar. In Victorian times the casks were placed in acellar but the beer was<br />
still drawn from taps into jugs<strong>and</strong>broughtupto the bar by`pot-boys'.The h<strong>and</strong> pulled<br />
pump, the beer engine, was developed in the early 20th century so that beer could be<br />
pumped up from the cellar to the bar. Beer engines were originally made of brass but<br />
are now almost always made of stainless steel or plastic to avoid any possibility of lead<br />
dissolvingintothebeer(Fig.23.3).Thesizeofthecylinderisusuallyaquarterorahalf<br />
imperialpint(150or300ml).Thetaponthecaskisconnectedtothecylinderbytubing<br />
which is usually aclear plastic (often PVC) <strong>and</strong> in this way yeast build up is easily<br />
seen.<br />
Typically, cask beers were dispensed at 10±15ëC (50±59ëF) <strong>and</strong> this was achieved by<br />
controlling cellar temperature. There has recently been adem<strong>and</strong> to drink cask beer at<br />
lower temperatures sometimes down to 7ëC (45ëF). Whether this is aresult of consumer<br />
dem<strong>and</strong> or has been led by marketing departments seeking to promote keg beer is<br />
uncertain! This has led to cooling cask beer en route from cask to bar. Using awatercooled<br />
cylinder in the beer engine, which usually means it will have the smaller volume<br />
of aquarter pint, can do this. Sometimes cask beer pythons are used (Fig. 23.4) in outlets<br />
serving a range of cask beers. There have been systems for introducing a gas mixture of<br />
carbon dioxide <strong>and</strong> nitrogen into the cask beer before dispense to increase the evenness of<br />
the foam <strong>and</strong> decrease bubble size (Watts, 2000). This is moving cask beer towards keg<br />
beer in sensory experience <strong>and</strong> would not find favour with beer traditionalists.<br />
The spout from the beer engine that delivers beer to the glass can be of various<br />
designs. It usually ends in a plastic removable tip, the sparkler, which can have a varying<br />
number of holes. Thus beer can be delivered through a `swan-neck' or straight spout to a<br />
tight or slack sparkler or even to no sparkler! This has a profound effect on the physical<br />
state of the beer in the glass, which can have a tight creamy head with much foam cling,<br />
or a slack open head with little foam cling as the beer is drunk. The variations are almost<br />
endless <strong>and</strong> are subject to much regional influence in the UK <strong>and</strong> much debate wherever<br />
cask beer drinkers meet.<br />
Hygiene<br />
Cleanliness is absolutely essential when dealing with cask beer. All the implements used<br />
must be scrupulously clean. Shives, pegs <strong>and</strong> keystones should be kept in a sealed<br />
container to avoid the pick-up of airborne moulds. There should be a permanent wall<br />
mounting for mallet, dipstick, taps <strong>and</strong> brushes. Cask beer delivery pipes must be cleaned<br />
at least weekly. They should be flushed with cold water <strong>and</strong> filled with detergent to the<br />
recommended concentration; an effective system would use 1% sodium hypochlorite<br />
solution. Soaking should be for 10±15 minutes <strong>and</strong> then a further 10±15 minutes with a<br />
fresh batch of detergent (Heron, 1992) this ensures all deposited <strong>and</strong> suspended materials<br />
are removed. Finally there must be a thorough rinse with cold water.<br />
Clean glasses are important. This applies equally to beer dispensed from kegs. Most<br />
bars use a glass washing machine <strong>and</strong> this must be correctly used <strong>and</strong> maintained. It must<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Outer wall of<br />
cylinder, made<br />
of brass, stainless<br />
steel or plastic<br />
Beer in<br />
Piston connected<br />
to h<strong>and</strong>le on bar<br />
Beer out<br />
Piston valve<br />
Cylinder<br />
containing beer<br />
Cylinder<br />
non-return<br />
valve<br />
Fig 23.3 Beer engine cylinder. This operates as asimple lift pump. When the h<strong>and</strong>le on the<br />
bar is pulled down the piston rises. The non-return valve at the beer inlet opens. Beer is drawn<br />
in beneath the piston. The piston valve remains closed. The h<strong>and</strong>le is then returned to the<br />
upright position. The cylinder valve now closes <strong>and</strong> the piston valve opens. The piston passes<br />
through the beer in the cylinder <strong>and</strong> the beer is lifted out of the cylinder to the bar counter <strong>and</strong><br />
dispensed. This is often through a`swan-neck' device (see Fig. 23.4). The process is repeated<br />
several times to deliver a full glass of beer at the bar.<br />
have adequate water pressure <strong>and</strong> flow rate <strong>and</strong> the detergent temperature must be<br />
accurately controlled. Replacement of the salt must be carried out according to the<br />
number of washing cycles actually used not simply on aweekly basis. Washed glasses<br />
must be left on aclean, odourless surface to cool completely before use. Glasses should<br />
be inspected by acompetent member of the bar staff to ensure they are absolutely clean<br />
<strong>and</strong> free of taint.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
E<br />
Empty casks<br />
Empty casks should be removed immediately from the stillage. The tap should be<br />
removed <strong>and</strong> thoroughly washed. Ahard peg should be driven into the shive <strong>and</strong> acork<br />
should be driven into the keystone hole. The cask should then be put in asuitable place<br />
for collection by the drayman.<br />
Throughput<br />
The final factor affecting the quality of cask beer is the throughput of the product. This is<br />
much morecritical than for the chilled <strong>and</strong> filtered kegbeer. The minimum throughput of<br />
any one quality should be two containers per week (Hunter, 1993) of whatever size is<br />
appropriatetotheoutlet.Acaskbeercontainerwhenbroachedshouldbesoldinnotmore<br />
than 72 hours <strong>and</strong> preferably within 48 hours. The frequent cause of poor cask beer<br />
quality is the failure to observe these guidelines.<br />
There is atrend in the UK to offer customers avery wide choice of cask beers in `real<br />
ale houses'. This has often resulted in the flouting of the throughput guidelines, with the<br />
result poor beer has been put on sale. This is not good for the image of cask beer <strong>and</strong> is a<br />
reason for the increase in sales ofthe nitrogenated keg ales (Section 23.4.1). There isalso<br />
thetrendtoservethecaskbeeratlowertemperaturesinanattempttoextenditsshelf-life.<br />
D<br />
O<br />
N<br />
M J<br />
C<br />
H<br />
B A<br />
Fig 23.4 Cask ale python to allow dispense of cask beer to temperatures 7ëC, often used in<br />
outlets serving a range of cask beers in low volume. A, stillage; B, cask tap; C, clear PVC<br />
tubing (0.375 in., 10 mm, internal diameter); D, pump; E, polythene tubing (0.375 in., 10 mm,<br />
external diameter); F, ale python; G, lager (main python); H, ale python driver; I, remote cooler;<br />
J, polythene tubing (0.375 in., 10 mm internal diameter); K, check valve; L, polythene tubing<br />
(0.375 in., 10 mm external diameter); M, Ü pint h<strong>and</strong>-pull with swan neck spout; N, agitator<br />
(sparkler) to suit product on dispense; O, pump clip for product on dispense (Lindsay, 2002).<br />
K L<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
I<br />
F<br />
G
These attempts are not always successful <strong>and</strong> the true flavour of the beer often suffers. At<br />
the present time the trend to keg beer continues <strong>and</strong> it seems likely that cask beer sales in<br />
the UK will continue to fall.<br />
23.4.3 Bottled <strong>and</strong> canned beer<br />
Drinking beer from bottles <strong>and</strong> cans is associated with drinking beer at home. Indeed in<br />
the UK off-trade sales of beer have risen from 10% of the total to over 30% of the total in<br />
the last 30 years (BLRA, 1999b). On-trade sales of small-pack beer were traditionally in<br />
returnable bottles but as a result of the influence of the package types sold in<br />
supermarkets, non-returnable bottles <strong>and</strong> even cans are now sold in the on-trade. Total<br />
sales of beer in cans in the UK have increased from 5% in 1974 to 25% in 2000 <strong>and</strong> in<br />
non-returnable bottles from < 1% in 1974 to 12% in 2000. Much of this beer is now<br />
drunk in pubs.<br />
International beer br<strong>and</strong>s are frequently sold in several package types, usually in<br />
bottles, cans <strong>and</strong> kegs. To preserve the identity <strong>and</strong> integrity of the br<strong>and</strong> it is essential<br />
that the quality is maintained in all package types. In the trade, therefore, it is important<br />
that beer served from a bottle or can is served correctly. This is not, of course, as difficult<br />
as with keg or cask beer. The main requirement is to serve clean bottles <strong>and</strong> to make sure<br />
the bottles are displayed attractively with the labels visible to the public. Almost always<br />
the bottle contents will be served cold, in some countries as low as 3 ëC (37 ëF). The<br />
bottles will be stored cold behind the bar in a refrigerator or in a cooled cabinet with clear<br />
glass doors.<br />
The pleasure of drinking some beers is enhanced by the use of br<strong>and</strong>ed glasses.<br />
However, there is a marked trend amongst young drinkers to drink from bottles when<br />
some of the pleasures of drinking beer are lost.<br />
23.5 Quality control<br />
Most brewing companies operate to the principles of quality assurance so that sound <strong>and</strong><br />
in-specification beer is delivered to their customers. The control of quality in the trade<br />
depends on the strict observance of all the principles of storage <strong>and</strong> dispense of beer<br />
discussed in this chapter. If these principles are followed then the customer should<br />
receive a perfect beer to drink.<br />
Many brewing companies employ inspectors to investigate premises where their beer<br />
is consumed. The focus is on the condition of the beer cellar, the storage of the beer <strong>and</strong><br />
on the condition of the beer dispense systems. Throughout the world these inspection<br />
systems have come under pressure as brewers have sought to save costs <strong>and</strong> less trade<br />
quality control work is now carried out. This is dangerous for the competitive position of<br />
beer related to other drinks.<br />
23.6 New developments in trade quality<br />
Conventional beer analyses do not always indicate how the beer will perform in the trade<br />
under the variety of conditions now prevalent (Watts, 2000). Foam quality <strong>and</strong> head<br />
retention on a beer, for example, has always been an important attribute. Head retention<br />
has been measured by a number of techniques amongst the most robust of which is the<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Rudin method (Rudin, 1958), which measures foam collapse under st<strong>and</strong>ard <strong>and</strong><br />
reproducible conditions. Recent studies have revealed beers with very similar Rudin<br />
values <strong>and</strong> yet having markedly different foam stabilities (Watts, 2000). This could<br />
representproblemsforabrewer.Foam stability isdefined asthe depthoffoamremaining<br />
on abeer after atwo-minute interval following ast<strong>and</strong>ard pour (Watts, 2000). A`visual<br />
assessment profile' has been developed to complement traditional beer analysis <strong>and</strong> so<br />
provide amore complete picture of the behaviour of the beer under trade conditions.<br />
Beers are assessed both quantitatively <strong>and</strong> qualitatively for foam quality, lacing,<br />
clarity, colour, `reflux' <strong>and</strong> `seeding' (see also Chapters 19 <strong>and</strong> 20). Reflux is a<br />
combination of the wave-like surge motion <strong>and</strong> overall bubble settling time on a draught<br />
dispensed beer. Seeding refers to the bubbles that rise up in the body of the beer. A<br />
trained panel of assessors examines each beer five times after a st<strong>and</strong>ard pour. This<br />
technique has been used to assess the effect of dispense head nozzle design on reflux. It<br />
was demonstrated that on a nitro-keg beer a nozzle pore size of 0.6 mm (0.024 in.) gave a<br />
significant increase in reflux compared to nozzle sizes of 0.7 mm (0.027 in.) <strong>and</strong> 0.8 mm<br />
(0.031 in.). This was ascribed to the greater shear when the beer was forced through the<br />
smaller nozzle. This provides considerable scope for changing the physical character of<br />
the beer where it is dispensed.<br />
Another factor of recent significance is the use of toughened glasses. There is<br />
considerably less seeding in lager beers poured into toughened glasses (Watts, 2000).<br />
This is presumably a result of the smoother surface not providing nucleation points for<br />
bubble formation.<br />
23.7 Summary<br />
Beer is drunk in a wide variety of premises, where the consumer forms his opinion of the<br />
beer. If the opinion is favourable then it is likely that the consumer will return <strong>and</strong> drink<br />
the beer again, as the brewer desires ± he seeks the delighted consumer. The brewer starts<br />
this process by assuring the highest quality of his beer to his customer, the vendor. It is<br />
then essential to ensure that the beer is stored <strong>and</strong> dispensed according to good <strong>practice</strong> at<br />
the premises where it is sold. This means scrupulous attention to the procedures of<br />
dispense for keg <strong>and</strong> cask beer, <strong>and</strong> detailed attention to hygiene of the cellar, dispense<br />
equipment <strong>and</strong> the glass.<br />
23.8 References<br />
BAMFORTH, C. W. (1998) Tap into the Art <strong>and</strong> Science of <strong>Brewing</strong>, New York, Insight Books.<br />
BLRA (1999a) Beer <strong>and</strong> Pub Facts, Brewers' <strong>and</strong> Licensed Retailers' Association, London.<br />
BLRA (1999b) Statistical H<strong>and</strong>book, Brewers' <strong>and</strong> Licensed Retailers' Association, London.<br />
BISHOP, L. R., WHITEAR, A. L. <strong>and</strong> INMAN, W. R. (1975) J. Inst. <strong>Brewing</strong>, 81, 131.<br />
HERON, P. C. (1992) Brewers' Guard. 121 (9), 29.<br />
HUNTER, A. R. (1993) The Brewer, 79, 159.<br />
LINDSAY, R. F. (2002) Personal communication.<br />
LINDSAY, R. F., LARSSON, E. <strong>and</strong> SMITH, I. B. (1996) Tech. Quart. MBAA, 33, 181.<br />
RUDIN, A. D. (1958) J. Inst. <strong>Brewing</strong>, 64, 238.<br />
STANLEY, J. M. (1999) Tech. Quart. MBAA. 36, 293.<br />
WATTS, E. (2000) Brewers' Guard. 129 (2), 20.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Appendix: units <strong>and</strong> some data of use in<br />
brewing<br />
Table A1 SI derived units<br />
Table A2 Prefixes for SI units<br />
Table A3 Comparison of thermometer scales<br />
Table A4 Interconversion factors for units of measurement<br />
Table A5 Specific gravity <strong>and</strong> extract table<br />
Table A6 Equivalence between Institute of <strong>Brewing</strong> units of hot water extract<br />
Table A7 Solution divisors of some sugars<br />
Table A8 Some properties of water<br />
Table A9 The density <strong>and</strong> viscosity of water at various temperatures<br />
Table A10 Some properties of water<br />
Table A11 The relationship between pressure <strong>and</strong> the temperature of steam<br />
Table A12 The solubility of pure gases in water at different temperatures<br />
Table A13 Salts in brewing liquors<br />
Table A14 Units of degrees of water hardness<br />
Table A15 Characteristics of some brewing materials<br />
Table A16 Pasteurization units (PUs)<br />
Fig. A1 The relationship between ethanol/water mixtures <strong>and</strong> the<br />
densities of the solutions.<br />
Introduction<br />
Although there is agradual trend towards the use of metric, or the derived SI units of<br />
measurement, their use is not universal <strong>and</strong>, of course, they were used less in the past.<br />
The nightmare muddle into which older British unit `systems' had degenerated is noted<br />
elsewhere,particularlyastheyrelatetocerealgrains<strong>and</strong>malt(Briggs,D.E.(1998)Malts<br />
<strong>and</strong> Malting, p. 742, London, Blackie Academic <strong>and</strong> Professional, or Aspen Publishers,<br />
Gaithersberg). American units, while having the same names as British units, often differ<br />
significantly in size. Thus, while the American gallon (gal. US) contains only 8pounds<br />
(lb.) of water or 8pints each weighing 1lb., the British gallon (gal. UK, or imp. gal.)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
contains 10lb., <strong>and</strong> so the British pint contains 1.25lb. of water. To have access to older<br />
<strong>and</strong> modern technical literature it is imperative that all units be understood <strong>and</strong> that<br />
readers have the information to allow them to convert reported values into units they are<br />
familiar with. In this book, wherever practical, metric or SI units have been used but<br />
equivalents in other units are given where this may be helpful even though this may be<br />
irritating to some readers, as it is to the authors.<br />
This appendix contains values for derived SI units <strong>and</strong> prefixes for them (Tables A1<br />
<strong>and</strong> A2), equivalents for thermometer scales (ëC, ëF <strong>and</strong> ëR; Table A3) <strong>and</strong><br />
interconversion factors for metric, British (imp. or UK) <strong>and</strong> American (US) units (Table<br />
A4). Table A5 gives the relationships between SG <strong>and</strong> extract values, while Table A6<br />
gives the approximate equivalents between hot water extracts as determined by the older<br />
<strong>and</strong> the modern methods of the Institute <strong>and</strong> Guild of <strong>Brewing</strong>. Table A7 gives some<br />
solution divisors for sugars. Various properties of water are given in Tables A8, A9 <strong>and</strong><br />
A10, including the pKw values <strong>and</strong> the amounts of oxygen dissolved from air or oxygen<br />
at different temperatures, the specific heats, latent heats, <strong>and</strong> surface tension at several<br />
temperatures. The relationship between steam pressure <strong>and</strong> temperature is summarized in<br />
Table A11. Table A12 gives the solubility of oxygen, nitrogen <strong>and</strong> carbon dioxide in<br />
water at different temperatures. Tables A13 <strong>and</strong> A14 give some data regarding salts in<br />
brewing liquors <strong>and</strong> units of water hardness. Table A15 contains information on the<br />
storage characteristics of some grist materials. Table A16 gives data on beer treatment<br />
temperatures <strong>and</strong> pasteurization units (PUs). Figure A.1 shows the relationship between<br />
the composition of ethanol/pure water, mixtures <strong>and</strong> their densities.<br />
As explained in the text, there are cases where there are no valid interconversion<br />
factors between analytical values. Thus the laboratory mashes specified in the<br />
Recommended Methods of the Institute <strong>and</strong> Guild of <strong>Brewing</strong> <strong>and</strong> Analytica-EBC<br />
(Section 1.15.1, p.9) aredifferent <strong>and</strong>can yielddifferent extract recoveries <strong>and</strong>yields of<br />
soluble nitrogen. In consequence there are no valid factors for interconverting HWE <strong>and</strong><br />
E, or the SNR <strong>and</strong> the Kolbach Index.<br />
Table A1 SI derived units. The base units are shown below<br />
Physical quantity Name Symbol Definition<br />
Energy joule J kg m 2 /s 2<br />
Force newton N kg m/s 2 ˆ J/m<br />
Power watt W kg m 2 /s 3 ˆ J/s<br />
Electrical charge coulomb C A s<br />
Electrical potential difference volt V kg m 2 /s 3 per A ˆ J/A per s<br />
Electrical resistance ohm kg m 2 /s 3 per A 2 ˆ V/A<br />
Inductance henry H kg m 2 /s 2 per A ˆ V/A per s<br />
Luminous flux lumen lm cd sr<br />
Illumination lux lx cd sr/m 2<br />
Frequency hertz Hz per s<br />
Pressure pascal Pa N/m 2<br />
Where m is the metre; kg is the kilogram; s is the second; A is the ampere; cd is the c<strong>and</strong>ela (luminous intensity);<br />
sr is the steridian (solid angle).<br />
Other base units are: Kelvin (K), thermodynamic temperature <strong>and</strong> temperature interval; mole (mol) molecular<br />
(or atomic) mass.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A2 Prefixes for SI units<br />
Fraction Prefix Symbol<br />
10 15<br />
10 12<br />
10 9<br />
10 6<br />
10 3<br />
10 2<br />
10 1<br />
10 1<br />
10 2<br />
10 3<br />
10 6<br />
10 9<br />
10 12<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
femto f<br />
pico p<br />
nano<br />
micro<br />
n<br />
milli m<br />
centi c<br />
deci d<br />
deka da<br />
hecto h<br />
kilo k<br />
mega M<br />
giga G<br />
tera T
Table A3 Comparison of thermometers showing the relative indications of the Fahrenheit,<br />
centigrade <strong>and</strong> ReÂaumur scales*<br />
Scale Boiling point Freezing point<br />
Fahrenheit (ëF) 212 32<br />
Centigrade (ëC) 100 0<br />
ReÂaumur (ëR)* 80 0<br />
Conversion of thermometer degrees<br />
1 ëC ˆ 1.8 ëF ˆ 0.8 ëR. ëC to ëR, multiply by 4 <strong>and</strong> divide by 5. ëC to ëF, multiply by 9. divide by 5,<br />
then add 32. ëR to ëC, multiply by 5 <strong>and</strong> divide by 4. ëR to ëF, multiply by 9, divide by 4, then add<br />
32. ëF to ëR, first subtract 32, then multiply by 4, <strong>and</strong> divide by 9. ëF to ëC, first subtract 32, then<br />
multiply by 5, <strong>and</strong> divide by 9.<br />
ëF ëC ëR ëF ëC ëR<br />
302 150 120 159.8 71 56.8<br />
284 140 112 158 70 56<br />
266 130 104 156.2 69 55.2<br />
257 125 100 154.4 68 54.4<br />
248 120 96 152.6 67 53.6<br />
239 115 92 150.8 66 52.8<br />
230 110 88 149 65 52<br />
221 105 84 147.2 64 51.2<br />
212 100 80 145.4 63 50.4<br />
210.2 99 79.2 143.6 62 49.6<br />
208.4 98 78.4 141.8 61 48.8<br />
206.6 97 77.6 140 60 48<br />
204.8 96 76.8 138.2 59 47.2<br />
203 95 76 136.4 58 46.4<br />
201.2 94 75.2 134.6 57 45.6<br />
199.4 93 74.4 132.8 56 44.8<br />
197.6 92 73.6 131 55 44<br />
195.8 91 72.8 129.2 54 43.2<br />
194 90 72 127.4 53 42.4<br />
192.2 89 71.2 125.6 52 41.6<br />
190.4 88 70.4 123.8 51 40.8<br />
188.6 87 69.6 122 50 40<br />
186.8 86 68.8 120.2 49 39.2<br />
185 85 68 118.4 48 38.4<br />
183.2 84 67.2 116.6 47 37.6<br />
181.4 83 66.4 114.8 46 36.8<br />
179.6 82 65.6 113 45 36<br />
177.8 81 64.8 111.2 44 35.2<br />
176 80 64 109.4 43 34.4<br />
174.2 79 63.2 107.6 42 33.6<br />
172.4 78 62.4 105.8 41 32.8<br />
170.6 77 61.6 104 40 32<br />
168.8 76 60.8 102.2 39 31.2<br />
167 75 60 100.4 38 30.4<br />
165.2 74 59.2 98.6 37 29.6<br />
163.4 73 58.4 96.8 36 28.8<br />
161.6 72 57.6 95 35 28<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A3 continued<br />
ëF ëC ëR ëF ëC ëR<br />
93.2 34 27.2 57.2 14 11.2<br />
91.4 33 26.4 55.4 13 10.4<br />
89.6 32 25.6 53.6 12 9.6<br />
87.8 31 24.8 51.8 11 8.8<br />
86 30 24 50 10 8<br />
84.2 29 23.2 48.2 9 7.2<br />
82.4 28 22.4 46.4 8 6.4<br />
80.6 27 21.6 44.6 7 5.6<br />
78.8 26 20.8 42.8 6 4.8<br />
77 25 20 41 5 4<br />
75.2 24 19.2 39.2 4 3.2<br />
73.4 23 18.4 37.4 3 2.4<br />
71.6 22 17.6 35.6 2 1.6<br />
69.8 21 16.8 33.8 1 0.8<br />
68 20 16 32 0 0<br />
66.2 19 15.2 30.2 1 0.8<br />
64.4 18 14.4 28.4 2 1.6<br />
62.6 17 13.6 23 5 4<br />
60.8 16 12.8 14 10 8<br />
59 15 12 4 20 16<br />
* In this table ëR is the ReÂaumur (not the Rankine) scale.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A4 Interconversion factors for units of measurement<br />
m ˆ 1.0936 yard ˆ 3.280 ft.; cm ˆ 0.39370 in.;<br />
hectare ˆ 2.471 acre; m 2 ˆ 10.764 ft. 2 ; cm 2 ˆ 0.1550 in. 2 ;<br />
m 3 ˆ 1000 dm 3 (or litre) ˆ 33.315 ft. 3 ˆ 61024 in. 3 ˆ 1.30795 yd. 3 ;<br />
hl ˆ 100 dm 3 (or litre) ˆ 21.998 gal. (British) ˆ 26.418 (US) ˆ 0.6111 barrel (British) ˆ 0.8387<br />
barrel (US) ˆ 0.8522 beer barrel (US) ˆ 0.1 m 3<br />
l ˆ 35.196 fl. oz. (British) ˆ 33.815 fl. oz. (US) ˆ 0.21998 gal. (British) ˆ 0.26418 gal. (US) ˆ<br />
0.035315 ft. 3<br />
tonne ˆ 1000 kg ˆ short ton ˆ 0.984207 long ton ˆ 2204.62 lb. ˆ 10 doppelzentner ˆ 20 zentner;<br />
zentner ˆ 50 kg ˆ 0.984 cwt ˆ 110,321 lb.; 1 kg ˆ 2.20462 lb.<br />
g ˆ 0.03527 oz. ˆ 15.432 grain; kg/m 3 ˆ g/dm 3 ˆ 0.062428 lb./ft. 3<br />
British measures<br />
yard ˆ 3 ft. ˆ 36 in. ˆ 0.9144 m; ft. ˆ 0.3048 m; lb./ft. 2 ˆ 4.88243 kg/m 2<br />
in. ˆ 2.540 cm ˆ 1000 thou (thous<strong>and</strong>th of an inch); lb./ft. 3 ˆ 16.0185 kg/m 3<br />
thou ˆ 25.4 micron or micrometer ( m);<br />
acre ˆ 4840 yd. 2 ˆ 0.4047 hectare (ha);<br />
yd. 2 ˆ 0.8361 m 2 ; ft. 2 ˆ 9.290 dm 2 ; in. 2 ˆ 6.4516 cm 2 ;<br />
yd. 3 ˆ 0.7646 m 3 ; ft. 3 ˆ 28.317 dm 3 ; in. 3 ˆ 16.3871 cm 3 ;<br />
ton (long) ˆ 20 cwt ˆ 2240 lb. ˆ 1016 kg ˆ 1.01605 tonne (short ton; metric);<br />
lb. ˆ 16 oz. ˆ 256 dram ˆ 7000 grains ˆ 0.45359 kg;<br />
oz. ˆ 28.35 g; grain ˆ 64.80 mg;<br />
gal. ˆ 160 fl. oz. ˆ 8 pints ˆ 1.201 gal. (US) ˆ 4.546 litre ˆ 0.1605 ft. 3 ˆ 0.125 bu (British);<br />
pint ˆ 0.5682 litre; fl. oz. ˆ 28.412 ml;<br />
butt ˆ 2 hogshead ˆ 3 barrel ˆ 108 gal. ˆ 4.9096 hl;<br />
brl ˆ 2 kilderkin ˆ 4 firkin ˆ 36 gal. ˆ 1.6365 hl ˆ 1.4 brl beer (US) ˆ 4.5 bu (British);<br />
bushel (bu, British) ˆ 8 gal. (UK) ˆ 36.3687 dm 3 ˆ 0.9690 bu (US); 8 bu ˆ 1 Qr.<br />
100% proof spirit (British) ˆ 57.10% ethyl alcohol (v/v) ˆ 49.28% ethyl alcohol (m/m), relative<br />
density 0.91976 at 60 ëF.<br />
US measures<br />
beer brl ˆ 31 gal. (US) ˆ 25.81 gal. (British) ˆ 1.734 hl ˆ 0.717 brl British;<br />
st<strong>and</strong>ard brl ˆ 31.5 gal. (US) ˆ 26.23 gal. (British) ˆ 1.1924 hl ˆ 0.729 brl (British);<br />
gal ˆ 8 pint ˆ 128 fl. oz. ˆ 3.7853 litre ˆ 0.8327 gal. (British) ˆ 231.0 in. 3<br />
bushel (bu, US) ˆ 35.23907 litre ˆ 1.03203 bu (British).<br />
Barley <strong>and</strong> malt measures<br />
Britain <strong>and</strong> South Africa: barley bushel ˆ 56 lb. ˆ 75.401 kg;<br />
barley quarter ˆ 448 lb. ˆ 4 cwt ˆ 0.2 ton ˆ 203.209 kg;<br />
malt bushel ˆ 42 lb. ˆ 19.051 kg;<br />
malt quarter ˆ 336 lb. ˆ 3 cwt ˆ 0.15 ton ˆ 152.407 kg<br />
Australia <strong>and</strong> New Zeal<strong>and</strong>: barley bushel ˆ 50 lb.; malt bushel ˆ 40 lb.<br />
US <strong>and</strong> Canada: barley bushel ˆ 48 lb.; malt bushel ˆ 34 lb.<br />
Useful data (some equivalents are approximate)<br />
kcal ˆ 4.186 kJ ˆ 3.968 btu (BTU) ˆ 1.1628 Wh ˆ 3088 ft. lb.<br />
btu ˆ 1.055 kJ ˆ 0.252 kcal ˆ 0.2931 Wh ˆ 778.2 ft. lb.<br />
Wh ˆ 3.6 kJ ˆ 0.860 kcal ˆ 3.412 btu ˆ 2655 ft. lb.<br />
therm ˆ 105.506 MJ ˆ 29.307 kWh;<br />
st<strong>and</strong>ard ton refrigeration per 24 h ˆ 12000 btu/h ˆ 3024 kcal/h<br />
atm ˆ 14.6959 lb./in. 2 ˆ 760 mm Hg (at 0 ëC <strong>and</strong> 45ë latitude) ˆ 101.325 kPa ˆ 1.01325 bar (ˆ<br />
33.899 ft. water ˆ 9.935 m of 1040 wort)<br />
lb./in. 2 ˆ 6896.76 N/m 2 ˆ 0.06895 bar ˆ 703 kg/m 3 ˆ 27.7 inches water<br />
lb./gal. (British) ˆ 99.76 g/l; lb./gal. (US) ˆ 119.8 g/l<br />
lb./brl (British) ˆ 3.336 g/l; lb./brl (US) ˆ 3.865 g/l<br />
grain/gal (British) ˆ 14.25 mg/l; grain/gal (US) ˆ 17.12 mg/l<br />
CO 2 in beer: g/100 ml ˆ 5.06 v/v in beer; v/v beer ˆ 0.198 g/100ml<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A5 Specific gravity <strong>and</strong> extract table<br />
The following table is based on those compiled by Dr Plato for the German Imperial Commission (Normal-Eichungskommission) <strong>and</strong> refers to the apparent<br />
specific gravities, as determined in the usual manner, by weighing in a specific gravity bottle in air or by means of a saccharometer. Cane sugar % w/v <strong>and</strong> % w/w<br />
represent g per 100 ml <strong>and</strong> g per 100 g of solution, respectively. The percentages by weight in column 6, corresponding with the specific gravities at 60 ëF given in<br />
column 1, were computed by interpolation from Plato's table for true specific gravities at 15ë/15 ëC <strong>and</strong> 16ë/15 ëC corrected to 60ë/60 ëF <strong>and</strong> then brought to 60 ëF/<br />
60 ëF in air by adding (SG ± 1) 0.00121. The cane sugar weight percentages were converted to volume percentages <strong>and</strong> the solution divisors calculated. The<br />
column headed Plato gives the specific gravities in air at 20 ë/20 ëC related to the cane sugar weight percentages <strong>and</strong>, with the latter, corresponds with the Plato<br />
table commonly used in breweries <strong>and</strong> laboratories where 20 ëC is the st<strong>and</strong>ard temperature. The column headed Balling similarly gives the specific gravities at<br />
17.5ë/17.5 ëC from the Balling table corresponding with the same sugar percentages. These specific gravities cannot accurately correspond with those at 60ë/60 ëF<br />
<strong>and</strong> 20ë/20 ëC on account of the errors in Balling's table. The following densities were used in the calculations:<br />
Water at 15 ëC/4 ëC 0.999126<br />
60 ëF/4 ëC 0.999035<br />
20 ëC/4 ëC 0.998234<br />
Specific gravity conversion table for cane sugar solutions<br />
British units Plato<br />
SG Brewers' Cane sugar Solution SG Cane sugar Balling BaumeÂ<br />
60 ëF pounds (% w/v) divisor 20 ëC (% w/w, Brix) SG 17.5 ëC Modulus 145<br />
1002.5 0.9 0.643 3.888 1.00250 0.641 1.00256 0.36<br />
1005.0 1.8 1.287 3.885 1.00499 1.281 1.00513 0.72<br />
1007.5 2.7 1.932 3.882 1.00748 1.918 1.00767 1.08<br />
1010.0 3.6 2.578 3.879 1.00998 2.552 1.01021 1.43<br />
1012.5 4.5 3.225 3.876 1.01247 3.185 1.01274 1.78<br />
1015.0 5.4 3.871 3.875 1.01496 3.814 1.01528 2.14<br />
1017.5 6.3 4.517 3.874 1.01745 4.439 1.01776 2.48<br />
1020.0 7.2 5.164 3.873 1.01993 5.063 1.02025 2.83<br />
1022.5 8.1 5.810 3.872 1.02242 5.682 1.02273 3.17<br />
1025.0 9.0 6.458 3.871 1.02490 6.300 1.02523 3.52<br />
1027.5 9.9 7.107 3.869 1.02740 6.917 1.02776 3.86<br />
1030.0 10.8 7.755 3.868 1.02989 7.529 1.03027 4.20<br />
1032.5 11.7 8.405 3.867 1.03238 8.140 1.03277 4.54<br />
1035.0 12.6 9.054 3.866 1.03486 8.748 1.03527 4.88<br />
1037.5 13.5 9.703 3.865 1.03736 9.352 1.03775 5.22<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
1040.0 14.4 10.354 3.863 1.03985 9.956 1.04024 5.55<br />
1042.5 15.3 11.003 3.862 1.04234 10.554 1.04273 5.88<br />
1045.0 16.2 11.652 3.862 1.04481 11.150 1.04523 6.21<br />
1047.5 17.1 12.303 3.861 1.04731 11.745 1.04773 6.54<br />
1050.0 18.0 12.953 3.860 1.04979 12.336 1.05022 6.87<br />
1052.5 18.9 13.604 3.859 1.05227 12.925 1.05269 7.20<br />
1055.0 19.8 14.255 3.858 1.05476 13.512 1.05515 7.52<br />
1057.5 20.7 14.907 3.857 1.05726 14.097 1.05760 7.84<br />
1060.0 21.6 15.560 3.856 1.05975 14.679 1.06005 8.16<br />
1062.5 22.5 16.213 3.855 1.06224 15.259 1.06252 8.48<br />
1065.0 23.4 16.866 3.854 1.06472 15.837 1.06500 8.80<br />
1067.5 24.3 17.519 3.853 1.06720 16.411 1.06747 9.12<br />
1070.0 25.2 18.173 3.852 1.06970 16.984 1.06995 9.44<br />
1072.5 26.1 18.827 3.851 1.07218 17.554 1.07244 9.75<br />
1075.0 27.0 19.482 3.850 1.07467 18.122 1.07494 10.06<br />
1077.5 27.9 20.135 3.849 1.07717 18.687 1.07743 10.37<br />
1080.0 28.8 20.791 3.848 1.07965 19.251 1.07990 10.69<br />
1082.5 29.7 21.446 3.847 1.08213 19.812 1.08237 11.00<br />
1085.0 30.6 22.101 3.846 1.08462 20.370 1.08486 11.30<br />
1087.5 31.5 22.758 3.845 1.08712 20.927 1.08737 11.61<br />
1090.0 32.4 23.414 3.844 1.08960 21.481 1.08986 11.91<br />
1092.5 33.3 24.071 3.843 1.09209 22.033 1.09235 12.21<br />
1095.0 34.2 24.726 3.842 1.09457 22.581 1.09481 12.51<br />
1097.5 35.1 25.384 3.841 1.09707 23.129 1.09730 12.81<br />
1100.0 36.0 26.041 3.840 1.09956 23.674 1.09980 13.11<br />
1102.5 36.9 26.700 3.839 1.10204 24.218 1.10230 13.41<br />
1105.0 37.8 27.360 3.838 1.10454 24.760 1.10480 13.71<br />
1107.5 38.7 28.019 3.837 1.10703 25.299 1.10730 14.00<br />
1110.0 39.6 28.679 3.836 1.10952 25.837 1.10983 14.30<br />
1112.5 40.5 29.339 3.834 1.11200 26.372 1.11235 14.59<br />
1115.0 41.4 30.000 3.833 1.11450 26.906 1.11486 14.88<br />
1117.5 42.3 30.660 3.832 1.11698 27.436 1.11735 15.17<br />
1120.0 43.2 31.321 3.831 1.11947 27.965 1.11984 15.46<br />
1122.5 44.1 31.981 3.830 1.12195 28.491 1.12231 15.74<br />
1125.0 45.0 32.643 3.829 1.12445 29.016 1.12478 16.03<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A5 Continued<br />
British units Plato<br />
SG Brewers' Cane sugar Solution SG Cane sugar Balling BaumeÂ<br />
60 ëF pounds (% w/v) divisor 20 ëC (% w/w, Brix) SG 17.5 ëC Modulus 145<br />
1127.5 45.9 33.305 3.828 1.12694 29.539 1.12729 16.31<br />
1130.0 46.8 33.970 3.827 1.12944 30.062 1.12980 16.60<br />
1132.5 47.7 34.632 3.826 1.13191 30.580 1.13228 16.88<br />
1135.0 48.6 35.295 3.825 1.13441 31.097 1.13477 17.16<br />
1137.5 49.5 35.958 3.824 1.13689 31.611 1.13723 17.44<br />
1140.0 50.4 36.621 3.823 1.13938 32.124 1.13971 17.72<br />
1142.5 51.3 37.285 3.822 1.14186 32.635 1.14221 18.00<br />
1145.0 52.2 37.951 3.821 1.14435 33.145 1.14473 18.27<br />
1147.5 53.1 38.617 3.820 1.14685 33.653 1.14727 18.55<br />
1150.0 54.0 39.284 3.818 1.14934 34.160 1.14980 18.82<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A6 Equivalence (approximate) between Institute of <strong>Brewing</strong> (UK) units of hot water extract<br />
Litre ë/kg (20 ëC) against lb/Qr (15.5 ëC, 60ëF) in the body of the table. Calculated from the Institute of <strong>Brewing</strong> Recommended Methods of Analysis (1993), on the<br />
basis that for lë/kg (20 ëC) the equivalent values in the lb./Qr will be 0.5 greater than those for lë/kg (15.5 ëC). Example: 251 lë/kg (20 ëC) is equivalent to 84.1 lb./Qr<br />
(15.5 ëC); 251 lë/kg (15.5 ëC) is equivalent to 83.6 lb/Qr (15.5 ëC).<br />
240 250 260 270 280 290 300 310 320 330<br />
0 80.2 83.7 87.2 90.8 94.2 97.8 101.3 104.8 108.3 111.9<br />
1 80.5 84.1 87.5 91.1 94.6 98.1 101.7 105.2 108.6 112.2<br />
2 80.8 84.4 87.9 91.4 94.9 98.5 102.0 105.5 109.0 112.6<br />
3 81.2 84.8 88.3 91.8 95.3 98.8 102.4 105.9 109.3 112.9<br />
4 81.5 85.1 88.6 92.1 95.7 99.2 102.7 106.2 109.7 113.2<br />
5 81.9 85.5 89.0 92.5 96.0 99.5 103.1 106.6 110.1 113.6<br />
6 82.3 85.8 89.4 92.8 96.4 99.9 103.4 106.9 110.4 113.9<br />
7 82.7 86.1 89.7 93.2 96.8 100.2 103.8 107.3 110.8 114.3<br />
8 83.0 86.5 90.1 93.5 97.1 100.6 104.1 107.6 111.2 114.6<br />
9 83.4 86.8 90.4 93.9 97.5 100.9 104.5 107.9 111.5 115.0<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A7 Solution divisors of some sugars at 20 ëC (68 ëF) (Pauls' Malt <strong>Brewing</strong> Room Book<br />
1995±7, p. 231). The apparent solids content of a solution (g/100 cm 3 ) = G/D where G is the excess<br />
specific gravity <strong>and</strong> D is the solution divisor<br />
SG (approx) D (sucrose) D (invert sugar) D (fructose) D (glucose) Wort solids<br />
1016.0 3.872 3.875 3.907 3.805 ±<br />
1020.0 3.869 3.872 3.904 3.805 ±<br />
1028.0 3.864 3.866 3.900 3.805 3.92<br />
1036.0 3.860 3.861 3.896 3.803 ±<br />
1061.0 3.849 3.845 3.882 3.794 3.90<br />
1083.0 3.840 3.832 3.869 3.784 ±<br />
1106.0 3.831 3.819 3.853 3.772 ±<br />
Table A8 Some properties of water at various temperatures (various sources)<br />
Temperature log Kw (pKw) Oxygen content of air-saturated water<br />
ëC ëF (mg/l) (mg/l) (cm 3 at NTP/kg)<br />
0 32 14.9435 ± 14.35 10.19<br />
5 41 ± ± 12.73 8.9<br />
10 50 14.5346 ± 11.25 7.9<br />
15 59 ± 9.8 10.06 7.0<br />
20 68 14.1669 8.8 9.08 6.4<br />
24 75.2 14.0000 ± ± ±<br />
25 77 13.9965 8.1 8.25 5.8<br />
30 86 13.8330 7.5 ± 5.3<br />
35 95 ± 7.0 ± ±<br />
50 122 13.2617 ± ± ±<br />
60 140 13.0171 ± ± ±<br />
Table A9 The density <strong>and</strong> viscosity ( ) at various temperatures of pure air-free water*<br />
Temperature<br />
ëC ëF (cP) Density (g/ml)<br />
10 14 ± 0.99812<br />
Ice, 0 32 ± 0.91700<br />
0 32 1.787 0.99987<br />
3.98 39.16 ± 1.00000<br />
5 41.00 ± 0.99999<br />
10 50.00 ± 0.99973<br />
15 59.00 ± 0.99913<br />
20 68.00 1.002 0.99823<br />
30 86.00 ± 0.99567<br />
50 122.00 0.5468 0.98807<br />
60 140.00 ± 0.98324<br />
65 149.00 ± 0.98059<br />
75 167.00 0.3781 ±<br />
80 176.00 ± 0.97183<br />
100 212.00 0.2818 0.95838<br />
*Data of Weast (1977), Moll (1979)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC
Table A10 Some properties of water (Bak et al., 2001; Moll, 1979)<br />
Temperature Specific heat Latent heat Latent heat Surface tension<br />
(cal/g/K) of vaporization of fusion (dynes/cm)<br />
(ëC) (ëF) (cal/g) (cal/g)<br />
6.62 20.1 ± ± 76 ±<br />
0 32 1.00738 594.8 79.7 (approx) 75.6<br />
10 50 1.00129 ± ± ±<br />
20 68 0.99883 ± ± 72.75<br />
100 212 ± 539 ± 58.90<br />
The specific thermal capacity of water is 4.19 kJ/kg/ëC, so the heat needed to raise the temperature of 1 m 3 of<br />
water from 20 ëC to 100 ëC is 335 MJ. The latent heat of vaporization of water is 2.26 MJ/kg, so the heat needed<br />
to evaporate of 1 m 3 water is 2260 MJ. Thus the total heat (energy) needed to vaporize of 1 m 3 water, initially at<br />
20 ëC, is 2595 MJ.<br />
Table A11 The relationship between the absolute pressure <strong>and</strong> the<br />
temperature of water-saturated steam<br />
Pressure<br />
(bar)<br />
0.5780<br />
0.7011<br />
0.8453<br />
1.000<br />
1.01325<br />
1.1<br />
1.2<br />
1.3<br />
1.5<br />
1.8<br />
2.0<br />
2.5<br />
3.0<br />
3.5<br />
4.0<br />
4.5<br />
5.0<br />
5.5<br />
6.0<br />
8.0<br />
10.0<br />
15.0<br />
20.0<br />
30.0<br />
40.0<br />
50.0<br />
Temperature<br />
(ëC) (ëF)<br />
85<br />
90<br />
95<br />
99.6<br />
100<br />
102.3<br />
104.8<br />
107.1<br />
111.4<br />
116.9<br />
120.2<br />
127.4<br />
133.5<br />
138.9<br />
143.6<br />
147.9<br />
151.8<br />
155.5<br />
158.8<br />
170.4<br />
179.9<br />
198.3<br />
212.4<br />
233.8<br />
250.3<br />
263.9<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
185<br />
194<br />
203<br />
211.3<br />
212<br />
216.1<br />
220.6<br />
224.8<br />
232.5<br />
242.4<br />
248.4<br />
261.3<br />
272.3<br />
282.0<br />
290.5<br />
298.2<br />
305.2<br />
311.9<br />
317.8<br />
338.7<br />
355.8<br />
388.9<br />
414.3<br />
452.8<br />
482.5<br />
507.0
Table A12 The solubility of pure gases in water under 1 atmosphere pressure <strong>and</strong> at the<br />
temperatures shown. The volumeof gasdissolved is given in ml,reduced to STP, that is at 0ëC <strong>and</strong><br />
1atmosphere pressure (760 mm of mercury at 0ëC, at latitude 45ë). (After Dawson et al., 1987)<br />
Temperature Gas solubility (ml at STP/litre water)<br />
(ëC) (ëF) O2 N2 CO2<br />
0<br />
5<br />
10<br />
15<br />
20<br />
25<br />
30<br />
35<br />
40<br />
45<br />
50<br />
32<br />
41<br />
50<br />
59<br />
68<br />
77<br />
86<br />
95<br />
104<br />
113<br />
122<br />
49<br />
43<br />
38<br />
34<br />
31<br />
28<br />
26<br />
24<br />
23<br />
22<br />
21<br />
24<br />
21<br />
19<br />
17<br />
15<br />
14<br />
13<br />
13<br />
12<br />
11<br />
11<br />
1713<br />
1424<br />
1194<br />
1019<br />
878<br />
759<br />
665<br />
592<br />
530<br />
479<br />
436<br />
The gram molecular volumeof an ideal gas at STP is 22.414 litres. The molecular weights of the gases are O2 =<br />
32.0; N2 =28.02; CO2 =44.01. Consequently 1ml each (non-ideal) gas, at STP, will contain approximately<br />
oxygen, 1.428 mg; nitrogen, 1.250 mg <strong>and</strong> carbon dioxide, 1.964 mg (compare with Table 10.5).<br />
Table A13 Salts in brewing liquors. Salt concentrations in brewing liquor are expressed in several<br />
different ways. This may be as the unit weight/unit volume, as in molarity (M, i.e. the gram<br />
molecular weight per litre) or as the equivalents per litre. Sometimes no account is taken of the fact<br />
that in dilute solution salts are nearly completely ionized, that is their component ions (cations,<br />
positively charged; anions, negatively charged) are separated in solution. Consequently it makes<br />
more sense to give the concentrations of the ions. At present these are usually expressed as mg (or<br />
g)/litre. In the past concentrations were often expressed as millvals, milligram equivalents/litre.<br />
When this is done the total concentration of the anions should equal the total concentration of the<br />
cations. The characteristics of some commonly encountered ions are indicated below. By ppm<br />
(parts/million) mg/litre are usually understood, but this is ambiguous <strong>and</strong> mg/kg may be intended<br />
Ion Symbol Atomic/molecular<br />
weight<br />
Calcium<br />
Magnesium<br />
Sodium<br />
Potassium<br />
Bicarbonate<br />
Carbonate<br />
Sulphate<br />
Chloride<br />
Nitrate<br />
Nitrite<br />
Ca 2+<br />
Mg 2+<br />
Na +<br />
K +<br />
HCO3<br />
CO 3 2<br />
SO4 2<br />
Cl<br />
NO3<br />
NO 2<br />
40.08<br />
24.32<br />
23.00<br />
39.10<br />
61.01<br />
60.01<br />
96.07<br />
35.48<br />
62.01<br />
46.01<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
Equivalent weight<br />
20.04<br />
12.16<br />
23.00<br />
39.10<br />
61.01<br />
30.01<br />
48.04<br />
35.48<br />
62.01<br />
46.01
Table A14 Units of degrees of water hardness. Equivalences between various units (degrees) of<br />
hardness in water. Total hardness usually means the sum of the calcium <strong>and</strong> magnesium expressed<br />
as equivalent amounts of CaO or CaCO3. This convention is historically interesting, but irrational,<br />
since calcium carbonate has limited solubility <strong>and</strong> calcium oxide (`quicklime') reacts violently with<br />
water to give calcium hydroxide (After Moll, 1979; Benson et al., 1997)<br />
French English German Ca Ca<br />
degree degree degree (mg/litre) (mM/litre)<br />
1 French degree a<br />
1.00 0.70 0.56 4.004 0.100<br />
1 English degree b<br />
1 German degree<br />
1.43 1.00 0.80 5.73 0.143<br />
c<br />
1 USA degree<br />
1.78 1.25 1.00 7.17 0.179<br />
d<br />
0.10 0.07 0.056 0.40 0.01<br />
1 Calcium (mg/l) 0.25 0.175 0.140 1.00 0.025<br />
Ca (millimoles/l) 0.00625 0.00438 0.0035 0.025 0.00063<br />
a CaCO3, 10 mg/l; b 1 Clark degree, CaCO3, 14.3 mg/l (1 grain/imp. gallon); c CaO, 10 mg/l;<br />
d CaCO3, 1 mg/l.<br />
Table A15 Some characteristics of raw materials <strong>and</strong> offal relevant to storage. The angle of<br />
repose is the angle from the horizontal adopted by materials when poured onto a flat surface. These<br />
values are somewhat variable depending on, for example, how well grain has been dressed, its<br />
moisture content <strong>and</strong> the variety. The valley angle is the included angle in the cone at the base of a<br />
silo or a store, which must not be exceeded if material is to run out freely (Briggs, 1998, Sugden et<br />
al., 1999).<br />
Material Bulk density<br />
(kg/m 3 )*<br />
Barley<br />
Malt<br />
Malt grist,<br />
roller-milled<br />
Malt grist,<br />
hammer milled<br />
Rice (polished)<br />
Rice (broken)<br />
Rye<br />
Sorghum<br />
Oats<br />
Wheat<br />
Maize grits<br />
Pelletized malt<br />
culms<br />
Malt culms<br />
Cereal dust<br />
610±618<br />
520±750<br />
510±550<br />
450±550<br />
300±370<br />
680±730<br />
780±850<br />
840±890<br />
680±740<br />
720±770<br />
350±520<br />
780±830<br />
630±700<br />
560±640<br />
224<br />
±<br />
Angle of repose<br />
(degrees)<br />
32<br />
23±35<br />
30±45<br />
26<br />
30-45<br />
±<br />
30<br />
30<br />
26<br />
30±45<br />
28±32<br />
30<br />
30±45<br />
±<br />
50<br />
60<br />
Maximum valley angle<br />
(degrees)<br />
*The reciprocals of these values indicate the amount of storage space needed for each kilogram of material (that<br />
is m 3 /kg)<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
116<br />
±<br />
90<br />
±<br />
90<br />
±<br />
120<br />
120<br />
128<br />
90<br />
116<br />
120<br />
90<br />
±<br />
±<br />
±
Table A16 Beer pasteurization; product temperature <strong>and</strong> lethal rate in Pasteurization Units, PUs.<br />
One PU is the lethal effect of holding for one minute at 60 ëC (140 ëF). Data selected from Paul's<br />
<strong>Brewing</strong> Room Book, 1998±2000, p. 271, where some practical considerations are considered<br />
Product temperature PU/min.<br />
(ëC) (ëF)<br />
53.0<br />
55.0<br />
56.5<br />
57.5<br />
59.0<br />
60.0<br />
61.5<br />
63.0<br />
64.0<br />
65.0<br />
66.5<br />
67.5<br />
69.0<br />
70.0<br />
71.5<br />
72.5<br />
74.0<br />
75.0<br />
76.5<br />
77.5<br />
79.0<br />
80.0<br />
127.4<br />
131.0<br />
133.7<br />
135.5<br />
138.2<br />
140.0<br />
142.7<br />
145.4<br />
147.2<br />
149.0<br />
151.7<br />
153.5<br />
156,2<br />
158.0<br />
160.7<br />
162.5<br />
165.2<br />
167.0<br />
169.7<br />
171.5<br />
174.2<br />
176.0<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC<br />
0.10<br />
0.19<br />
0.32<br />
0.45<br />
0.72<br />
1.0<br />
1.7<br />
2.7<br />
3.7<br />
5.2<br />
8.6<br />
12<br />
19<br />
27<br />
45<br />
62<br />
100<br />
139<br />
231<br />
320<br />
519<br />
720
Ethanol (%)<br />
13<br />
12<br />
11<br />
10<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
Composition by<br />
weight (in vacuo)<br />
Composition by volume<br />
(in air at 20°C)<br />
Pure ethanol has a density,<br />
in air at 20°C, of 788.16 kg/m 3 .<br />
Under these conditions water has<br />
a density of 997.15 kg/m 3<br />
0 980 1 2 3 4 5 6 7 8 9 990 1 2 3 4 5 6 7 998<br />
Density (kg/m 3 ) in air at 20°C<br />
Fig. A1 The relationships between the densities (kg/m 3 ) of alcohol (ethanol)/water mixtures, by<br />
volume in air at 20 ëC, or by mass in vacuo. Data from the <strong>Brewing</strong> Room Book, 1998±2000, p. 281,<br />
Pauls Malt, Ipswich <strong>and</strong> Kentford, Suffolk, UK.<br />
References<br />
BAK, S. N., EKENGREN, OÈ ., EKSTAM, K., HAÈ RNULV, G., PAJUNEN, E., PRUCHA, P. <strong>and</strong> RASI, J. (2001) Water in<br />
<strong>Brewing</strong>: a European Brewery Convention Manual of Good Practice. 128 pp. NuÈrnberg.<br />
Fachverlag Hans Carl.<br />
BENSON, J. T., COLEMAN, A. R., DUE, J. E. B., HENHAM, A. W., TWAALFLLOVEN, J. G. P. <strong>and</strong> VINCKX, W. (1997)<br />
Brewery Utilities: a European Brewery Convention Manual of Good Practice. 200 pp. NuÈrnberg.<br />
Fachverlag Hans Carl.<br />
BRIGGS, D. E. (1998) Malts <strong>and</strong> Malting. p. 750. London: Blackie Academic & Professional. Gaithersburg:<br />
Aspen Publishing.<br />
DAWSON, R. M. C., ELLIOTT, D. C., ELLIOTT, W. H. <strong>and</strong> JONES, K. M. (1987) Data for Biochemical Research.<br />
(3rd edn). 580 pp. Oxford. The Clarendon Press.<br />
MOLL, M. (1979) in <strong>Brewing</strong> Science, 1, (J. R. A. Pollock, ed.) p. l. London. Academic Press.<br />
SUGDEN, T. D., WEBB, C., BYRNE, H., VAN WAESBERGHE, J. <strong>and</strong> WULFF, T. (1999) Milling: a European<br />
Brewery Convention Manual of Good Practice. 102 pp. NuÈrnberg. Fachverlag Hans Carl.<br />
WEAST, R. C. (ed.) (1977) CRC H<strong>and</strong>book of Chemistry <strong>and</strong> Physics. (58th edn). Clevel<strong>and</strong>. CRC Press.<br />
Copyright © 2004 Woodhead Publishing Limited <strong>and</strong> CRC Press, LLC